Return-Path: X-Original-To: archive-asf-public-internal@cust-asf2.ponee.io Delivered-To: archive-asf-public-internal@cust-asf2.ponee.io Received: from cust-asf.ponee.io (cust-asf.ponee.io [163.172.22.183]) by cust-asf2.ponee.io (Postfix) with ESMTP id 4A34A200C05 for ; Mon, 23 Jan 2017 23:34:15 +0100 (CET) Received: by cust-asf.ponee.io (Postfix) id 487EB160B49; Mon, 23 Jan 2017 22:34:15 +0000 (UTC) Delivered-To: archive-asf-public@cust-asf.ponee.io Received: from mail.apache.org (hermes.apache.org [140.211.11.3]) by cust-asf.ponee.io (Postfix) with SMTP id DAFCA160B3C for ; Mon, 23 Jan 2017 23:34:10 +0100 (CET) Received: (qmail 63617 invoked by uid 500); 23 Jan 2017 22:34:09 -0000 Mailing-List: contact dev-help@kafka.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Reply-To: dev@kafka.apache.org Delivered-To: mailing list dev@kafka.apache.org Received: (qmail 63605 invoked by uid 99); 23 Jan 2017 22:34:09 -0000 Received: from pnap-us-west-generic-nat.apache.org (HELO spamd2-us-west.apache.org) (209.188.14.142) by apache.org (qpsmtpd/0.29) with ESMTP; Mon, 23 Jan 2017 22:34:09 +0000 Received: from localhost (localhost [127.0.0.1]) by spamd2-us-west.apache.org (ASF Mail Server at spamd2-us-west.apache.org) with ESMTP id BA2A11A040F for ; Mon, 23 Jan 2017 22:34:08 +0000 (UTC) X-Virus-Scanned: Debian amavisd-new at spamd2-us-west.apache.org X-Spam-Flag: NO X-Spam-Score: 0.679 X-Spam-Level: X-Spam-Status: No, score=0.679 tagged_above=-999 required=6.31 tests=[DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, KAM_ASCII_DIVIDERS=0.8, RCVD_IN_DNSWL_NONE=-0.0001, RCVD_IN_MSPIKE_H3=-0.01, RCVD_IN_MSPIKE_WL=-0.01, SPF_PASS=-0.001] autolearn=disabled Authentication-Results: spamd2-us-west.apache.org (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com Received: from mx1-lw-eu.apache.org ([10.40.0.8]) by localhost (spamd2-us-west.apache.org [10.40.0.9]) (amavisd-new, port 10024) with ESMTP id KWzywo3EvWAd for ; Mon, 23 Jan 2017 22:33:40 +0000 (UTC) Received: from mail-ot0-f170.google.com (mail-ot0-f170.google.com [74.125.82.170]) by mx1-lw-eu.apache.org (ASF Mail Server at mx1-lw-eu.apache.org) with ESMTPS id A57435F23A for ; Mon, 23 Jan 2017 22:33:39 +0000 (UTC) Received: by mail-ot0-f170.google.com with SMTP id 73so114122642otj.0 for ; Mon, 23 Jan 2017 14:33:39 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :content-transfer-encoding; bh=++idiyI+Qw1fCEISUlOrtmchYEIkD4d1zTz10PuDk1I=; b=QMmDyn3tppwor+OEUkF1zFM0v3X+HUQAIsDIbr6PAkCrJRK8LX8q91gMcVHL2DEguq 8/eEIoMSIWyxJgn8z+ibL3ZWbU0/pjrXSEydHyRZ9ru7Ku0bXgJMqPcCeFFGnpl7zeiE TDn9jthv/bLFXdCIz95ZarV8md6CEnv/tadRj4+uU+yBnXoQ23w90529a845jDfe9pEI m+g4Qo6AJYmbkeMPR4CutshLn3RNWAQSyzfsPIpzlfbnX1E5mECg1aoCk+rACAEGNXhR CZ5E4rmneLjD5mvm5Fw5YvtP9iDrzWrnB7Q3zEyvwYJvBNnqRxMfvkHsUHqjJ7bZlb+C V2UA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:content-transfer-encoding; bh=++idiyI+Qw1fCEISUlOrtmchYEIkD4d1zTz10PuDk1I=; b=an9OzJR9Mz34DbcIRi/CrkEn4buLVosqJQzFq/Gyn4Pfxb7N5SPdQXZ8WvaairnrxD i74sYzjKBOG/Q88nkNGwZn3Yexoh+Yw6vk5+hBXaGs0COIgLQccpJ22jmRVdZH6EixGB GiDLOTNISMbpYFu17qcz0lIVnbftZXGGXdvI54xuXyY/Y7020wRmKlksV1xOQthBYNdq HeGzwU+vw2AOn4Wo+LfAJkj/hp3TM4OBaxkjVfGgJhNkAvgu/csgkDmflN8J6IDNdiCQ 3tI9EgR0twBFRvmkh5jQ9/VaiW7X6gHRtNQeYDmVAiKzmHhzz0V1EVo7DWvZYIRtzZAQ CZUg== X-Gm-Message-State: AIkVDXJ95X6sMVAxPqJOFmJFMu8mLFedsd/uLdUZiVw3KVxzc0l2Q3zI7W8Nf8bFDU8Rxx2KbECFtAH1odbnvA== X-Received: by 10.157.9.69 with SMTP id 63mr16783842otp.152.1485210817503; Mon, 23 Jan 2017 14:33:37 -0800 (PST) MIME-Version: 1.0 Received: by 10.157.56.167 with HTTP; Mon, 23 Jan 2017 14:33:36 -0800 (PST) In-Reply-To: References: <5E35A146-E6AB-4D7E-8E64-B79D9B9517AE@ig.com> <6427404E-1CD0-49D5-A640-9649A7358B9B@ig.com> <922acbb8-5a5e-9cab-e03f-e335a68c8698@confluent.io> <4aee81ed-0afa-ccb6-a583-b9f2bd95e4aa@confluent.io> <1482214102094.20103@ig.com> <716FC82D-FEC3-4C42-B1D9-3AE6B96C6792@ig.com> From: Sean McCauliff Date: Mon, 23 Jan 2017 14:33:36 -0800 Message-ID: Subject: Re: [DISCUSS] KIP-82 - Add Record Headers To: dev@kafka.apache.org Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable archived-at: Mon, 23 Jan 2017 22:34:15 -0000 Scroll down to "Parse Header, create collection, populate collection search collection" https://docs.google.com/document/d/1tfT-6SZdnKOLyWGDH82kS30PnUkmgb7nPLdw6p6= 5pAI/edit#heading=3Dh.66o3tyythvas If you want string keys this means accepting a header deserialization time of 1.5-2.5x slower than having int keys. It's not super clear what performance target one wants to hit but if the user needs to access the header keys then header processing would be the limiting CPU performance factor when 1M or so messages needs to be processed per second and there are header 5-20 keys that need to be accessed. This would be much lower if the accessing code actually needs to do something the associated value. If performance is really a concern then int keys make the most sense. Compression will probably not make CPU utilization better. Really only a few people in any organization should care about the value of these keys and that number may actually be zero if there is a large third party community that provides solutions that uses the headers. The only issue is then key namespace which is much better if people could manage their own namespaces with strings. Otherwise some kind of namespace scheme with integers (or longs) needs to be created as others have pointed out. Preregistering headers means that all producers and consumers need to see the same mapping. This seems very problematic to get right with a large Kafka deployment. The lazy map thing is a good idea but is really a micro optimization. This can be added later. Rather than having a new type, like Headers , which involves more object allocation, just have a method like byte[] header(int headerKey) in ConsumerRecord that completely hides the implementation of headers. Regarding the need for detailed use cases for headers: the use case for headers is not mixing control (or metadata) with data; that's it. If I have metadata in my data then it's difficult to reuse code that wants this data. If I have data in my metadata then it's difficult to reuse the control code that consumes the metadata. This is a basic concept that is is implemented in many protocols, serialized data formats, file systems and more. Sean On Wed, Jan 11, 2017 at 9:05 AM, Jay Kreps wrote: > Yeah I think I share a couple of concerns around the details of the > proposal beyond the use cases: > > 1. I think numeric keys are just a non-starter for a user-facing > interface that is meant to work across third party components (not jus= t > within an organization). We've gone down this path of making cryptic > interfaces before based on what's easy to implement (simple consumer!)= . The > result is a mess we are still trying to walk away from. I think we are > solving a compression/normalization problem by having the user impleme= nt a > compression algorithm in their head and share it across their company = and > between components and plugin. This is definitely easier for us, but n= ot > the right interface! I think we should consider instead (1) just take = the > perf hit and claim it's okay since it's optional (I don't necessarily = buy > that since it makes a lot of common use cases pretty questionable), (2= ) > work on getting more efficient compression in place so we can recommen= d > that without reservation (e.g. if that facebook lib pans out) or (3) a= llow > the client to optionally pre-register headers to get a more compact > numerical mapping. > 2. I think we should think about creating the Map lazily to avoid > parsing out all the headers into little objects. HashMaps themselves a= re > kind of expensive and the consumer is very perf sensitive so and makin= g > gazillions of hashmaps that may or may not get used is probably a bad = idea. > 3. We should think about having our own type, Headers or something > rather than a Map. I kind of dislike this kind of unnecessary type in > internal code, but for this kind of public API the ability to get just= the > methods that make sense and document the details of the semantic can b= e > worthwhile. > > -Jay > > On Tue, Jan 10, 2017 at 10:38 PM, Ewen Cheslack-Postava > wrote: > >> I'm horribly behind and outdated on this thread, but wanted to mention o= ne >> other potential issue. >> >> The KIP talks about using simple Maps for the headers. For a lot of this >> discussion, I tend to either think of completely opaque headers (i.e jus= t >> another byte[] blob, which seems to have been rejected based on >> interoperability grounds but I think has the nice property of making it >> easy to blame the implementer of the serde for any problems with the hea= der >> implementation) or HTTP-style headers (seemingly the most ubiquitous >> example of headers, though certainly there are many others, and one whic= h >> we/users may need/want to interoperate with). >> >> Strictly speaking, the keys on HTTP headers aren't unique. See >> https://www.w3.org/Protocols/rfc2616/rfc2616-sec4.html#sec4.2 >> specifically: >> >> > Multiple message-header fields with the same field-name MAY be present= in >> a message if and only if the entire field-value for that header field is >> defined as a comma-separated list [i.e., #(values)]. It MUST be possible= to >> combine the multiple header fields into one "field-name: field-value" pa= ir, >> without changing the semantics of the message, by appending each subsequ= ent >> field-value to the first, each separated by a comma. The order in which >> header fields with the same field-name are received is therefore >> significant to the interpretation of the combined field value, and thus = a >> proxy MUST NOT change the order of these field values when a message is >> forwarded. >> >> The spec guarantees that you can collapse repeated headers into a single >> instance. But in practice there's a ton of stuff that doesn't. And every >> robust HTTP library I've encountered uses a multimap to represent header= s. >> >> I'm not necessarily advocating for supporting this, just suggesting that= we >> think about it. As you add these features and get closer to mapping to >> other systems, people will inevitably try to map them. Headers are an ar= ea >> where, if we're going to add them, it's worth considering compatibility = as >> someone will inevitably come and complain that system X does Y with head= ers >> and we should also support Y because any decent system that provides >> headers will do so. >> >> -Ewen >> >> On Tue, Jan 10, 2017 at 5:27 PM, Jay Kreps wrote: >> >> > Hey Gwen, yes, your approach is much better than what I described and >> > solves one of my objections. I didn't think of doing it in MM rather t= han >> > in the original producer. I don't think it adds any overhead beyond th= e >> > additional header bytes if I'm not mistaken, since MM already unpacks = and >> > repacks messages (unlike the broker which tries to avoid repacking and >> > hence can't efficiently add a header). I think my other objection arou= nd >> > size is likely solvable too by some kind of aliasing scheme? If you gu= ys >> > think this can be made to work then I think we have two pretty credibl= e >> use >> > cases with this and the tracing id family of uses, plus a number of >> things >> > that are at least moderately improved. If so, I'm convinced. >> > >> > -Jay >> > >> > >> > On Tue, Jan 10, 2017 at 8:48 AM, Michael Pearce >> > wrote: >> > >> > > Hi Jay et al. >> > > >> > > On the replication using cluster id. >> > > >> > > Two options depending on how a company wants or needs to implement >> their >> > > replication as there are options of doing this as a mesh, star, or >> ring. >> > > >> > > Either as some companies enforce a provider/custom producer/consumer >> > > wrapper, it can be enforced the interceptors needed to be added, as = you >> > can >> > > enforce to get support of the platform you must use company x=E2=80= =99s >> internal >> > > wrappers. >> > > An alternative approach is to do this at the mirror maker, as Gwen >> noted >> > > we can actually make the mirror maker append this, like wise for a m= ore >> > > complex implementation, where you can have multiple hops/routes this >> can >> > > also be the message routing. >> > > >> > > Re the size of the clusterId, yes the auto generated one is =E2=80= =9Ccough=E2=80=9D >> > > hideous in size. >> > > >> > > Most organisations though will only need a space to uniquely identif= y >> 127 >> > > (byte) clusters, or at worst 2 bytes (short) they need to mirror and >> > > replicate between, and as such can either as platform teams generall= y >> > will >> > > roll out the mirror makers / replicators with the cluster can either >> > > manually adjust the clusterId on ZK, or use a mapping file all their >> > mirror >> > > makers have to map the larger id to a smaller one. >> > > >> > > Also as Gwen states this is something you should be able to opt into= or >> > > not, this is the benefit of headers, if you don=E2=80=99t use it, it= doesn=E2=80=99t >> use >> > up >> > > the space, its entirely optional. >> > > >> > > Regards >> > > Mike >> > > >> > > >> > > On 09/01/2017, 19:40, "Gwen Shapira" wrote: >> > > >> > > Mirroring + ClusterIDs: >> > > >> > > Kafka clusters already have an automatically generated ID that i= s >> > > exposed to clients: >> > > https://cwiki.apache.org/confluence/display/KAFKA/KIP- >> > 78%3A+Cluster+Id >> > > >> > > It sounds trivial to get MirrorMaker's consumer to grab this ID >> > (which >> > > always exists) from the source cluster and add it to the message >> > > headers. >> > > >> > > Obviously adding stuff to the message makes it larger so this >> should >> > > be a feature you should be able to turn on/off depending on the >> > > replication requirements. If you need the feature - you pay the >> > price. >> > > I don't think there is an immediate need to put an effort into >> > > shrinking the Cluster ID, but others can comment on that. >> > > >> > > >> > > >> > > >> > > >> > > On Mon, Jan 9, 2017 at 11:19 AM, Jay Kreps >> wrote: >> > > > Hey Michael, >> > > > >> > > > Sorry for lagging, I've been out for the holiday break. What >> would >> > > help me >> > > > would be to get to a consolidated list of use cases we've work= ed >> > out >> > > and >> > > > think add real value. I think getting down to the details real= ly >> > > matters >> > > > for these. >> > > > >> > > > The two lists I'm aware of are the one in the KIP and the one = you >> > > gave in >> > > > this email thread. Both seem to be more at the sketch level. L= et >> me >> > > review >> > > > the use cases and see if I have it right. If so let's get it i= n >> the >> > > design >> > > > proposal. >> > > > >> > > > Your use cases: >> > > > >> > > > 1. Tracing IDs -- I get this use case and think I see how i= t >> > > would work. >> > > > I think because the tracing is from a third party system th= ere >> > is >> > > some >> > > > value to having a header versus a custom format to make thi= s >> > work >> > > fully >> > > > transparently. >> > > > 2. Mirroring -- I'm a bit unsure how this will work, I'll >> > explain >> > > my >> > > > thinking below >> > > > 3. End-to-end encryption -- I think what you are saying is >> that >> > > if you >> > > > want "routing' fields that are accessible in transit withou= t >> > > decrypting you >> > > > need a scheme to leave this decrypted.I think the idea is t= o >> > avoid >> > > > decrypting to get a header, but I'm not sure if I get how t= his >> > is >> > > better >> > > > than a custom message format that doesn't encrypt the thing= s >> you >> > > don't want >> > > > encrypted since you have to be aware of the scheme anyway. >> > > > 4. EOS -- I'm not convinced this is a good approach for EOS= . >> > > > >> > > > Linked from KIP: >> > > > >> > > > 1. Schema-Registry-aware avro serialization -- not clear ho= w >> > this >> > > is >> > > > better than keeping the schema id in the value >> > > > 2. Polyglot kafka topics -- marginally better than a contai= ner >> > > format, >> > > > but of course you still have to all agree on the the header >> and >> > > > header=3D>serialization mapping so not sure if it's really >> better >> > > than a >> > > > container format. Also not wild about polyglot topics :-) >> > > > 3. End-to-End routing trail including mirroring --- I'd put >> this >> > > under >> > > > tracing ids >> > > > 4. Cross-service process tracing -- ditto >> > > > 5. Signing -- I like this use case but it seems to depend o= n a >> > > feature >> > > > we don't have and aren't proposing adding, right? >> > > > 6. Per-record custom compression -- why do you want this? >> > > compressing >> > > > per message won't work well compared to batch compression, >> > right? >> > > > 7. Large Message Support -- I think we can agree that altho= ugh >> > > this >> > > > might work it's clearly not the right way to support large >> > > messages in Kafka >> > > > >> > > > Matthias >> > > > >> > > > 1. Control messages --- Matthias, not sure if I get the >> > > motivation in >> > > > streams for this. If we did want to do it, wouldn't it be >> easier >> > > to >> > > > implement and by allowing users to send control messages (i= n >> the >> > > EOS sense) >> > > > and adding a consumer config to allow either passing these = on >> or >> > > > suppressing them? >> > > > >> > > > So I if I were to summarize I'd list the use cases as follows: >> > > > >> > > > - Much better with headers: tracing IDs >> > > > - Somewhat better: end-to-end encryption, polyglot topics >> > > > - Don't buy: Mirroring (though perhaps could be made to wor= k >> > with >> > > > changes to the proposal), EOS, large messages, compression >> > > > - Not sure: Control messages >> > > > >> > > > Does this match where other people are at? From my point of vi= ew >> > for >> > > > something like this where we are changing the core data model = we >> > > should >> > > > ideally get a handful of use cases to the point where they are >> > > really well >> > > > thought out and we know they can use the feature. I think we h= ave >> > > one solid >> > > > one with tracing IDs, and I think maybe mirroring could be >> another >> > > with >> > > > some work. I'm less sure about the others. >> > > > >> > > > Bi-directional Mirroring >> > > > I think bidirectional mirroring could potentially become a >> > > compelling use >> > > > case if we an find a way to make it work. If I understand it, = the >> > > proposal >> > > > is to add a header like "origin-cluster" with the cluster id o= f >> the >> > > Kafka >> > > > cluster. Then modify the mirroring tool to disallow mirroring = a >> > > message >> > > > back to it's own origin cluster. This basic idea is definitely >> > > workable and >> > > > useful, but can we really get to a good implementation using >> > headers? >> > > > I imagine the approach would be to add this header in the >> producer >> > > if some >> > > > config is set. There are two problems: >> > > > >> > > > - You add a pretty big honking cluster ID to each message. = For >> > > small >> > > > messages this maybe larger than the contents itself. >> > > > - What happens if someone somewhere forgets to set that con= fig >> > on >> > > some >> > > > client somewhere? >> > > > >> > > > If someone forgets to set the config in some client somewhere = I >> > > guess you >> > > > have various options in mirror maker: >> > > > >> > > > 1. drop the data (pretty bad, now your data is out of sync) >> > > > 2. crash (pretty bad--how do you recover) >> > > > 3. insert some default (pretty bad since would also lead to >> data >> > > loss if >> > > > that's not right) >> > > > >> > > > Basically although I think a cluster tag could be a good way t= o >> do >> > > > bi-directional mirroring, I'm not sure this naive approach I >> > > described >> > > > would get us to a usable implementation. This could be fixable= . I >> > > think if >> > > > there was some scheme to check required headers on the server >> that >> > > could >> > > > alleviate the missing header problem. Not really sure what to = do >> > > about the >> > > > message bloat problem, but maybe there could be some way to >> shrink >> > > the >> > > > cluster id? >> > > > >> > > > Cheers, >> > > > >> > > > -Jay >> > > > >> > > > >> > > > On Mon, Dec 19, 2016 at 10:08 PM, Michael Pearce < >> > > Michael.Pearce@ig.com> >> > > > wrote: >> > > > >> > > >> @Jay, Ismael and a few others. >> > > >> >> > > >> Based on the discussion on kip-87, and the additional use cas= e >> by >> > > Matthias. >> > > >> >> > > >> Just for clarity do we now have consensus on the earlier poin= ts >> > (1) >> > > >> headers are useful (2) headers should be native in Kafka? >> > > >> >> > > >> Cheers >> > > >> Mike >> > > >> ________________________________________ >> > > >> From: radai >> > > >> Sent: Monday, December 19, 2016 5:01 PM >> > > >> To: dev@kafka.apache.org >> > > >> Subject: Re: [DISCUSS] KIP-82 - Add Record Headers >> > > >> >> > > >> having said that im perfectly fine with something like a "no >> > > payload" flag >> > > >> (as opposed to null payload), which would mean such a msg is = not >> > > returned >> > > >> from poll() unless the user code specifically "opts in" and >> would >> > > serve >> > > >> only to carry headers (making it a control message) >> > > >> >> > > >> On Mon, Dec 19, 2016 at 8:58 AM, radai < >> > radai.rosenblatt@gmail.com> >> > > wrote: >> > > >> >> > > >> > if all you want is to tag "end of stream" wouldnt a header = on >> > the >> > > last >> > > >> msg >> > > >> > in the stream be enough? >> > > >> > >> > > >> > On Sun, Dec 18, 2016 at 2:03 PM, Matthias J. Sax < >> > > matthias@confluent.io> >> > > >> > wrote: >> > > >> > >> > > >> >> The details about headers for control messages are still t= o >> > > define. But >> > > >> >> yes, the idea is to have some common default behavior that >> > > clients would >> > > >> >> need to implement. >> > > >> >> >> > > >> >> The point is, that "regular headers" add meta data to regu= lar >> > > messages. >> > > >> >> Thus, those messages will be returned to the user via >> .poll(). >> > > And after >> > > >> >> the message is received the user can check if meta data is >> > > present and >> > > >> >> read it. >> > > >> >> >> > > >> >> For control messages, we do not want those to pop up via >> > .poll() >> > > as >> > > >> >> those are no regular messages. A client would need to opt-= in >> to >> > > see >> > > >> >> those messages (either via poll() or maybe a callback). Th= us, >> > we >> > > need >> > > >> >> some special (standardized) header IDs that indicate contr= ol >> > > messages >> > > >> >> that should not be returned to the user via poll() by >> default. >> > > >> >> >> > > >> >> >> > > >> >> -Matthias >> > > >> >> >> > > >> >> On 12/17/16 9:37 PM, Roger Hoover wrote: >> > > >> >> > Matthias, >> > > >> >> > >> > > >> >> > Thanks for your input. I'm +1 on control messages as th= ey >> > > seem to be >> > > >> >> the >> > > >> >> > simplest way to implement watermarks ( >> > > >> >> > https://www.oreilly.com/ideas/the-world-beyond-batch- >> > > streaming-102), >> > > >> a >> > > >> >> > feature that would add a lot of value to Kafka Streams >> IMHO. >> > > >> >> > >> > > >> >> > Your argument that the control-message indicator belongs= in >> > the >> > > >> >> client-only >> > > >> >> > section of the record format make sense. Just to make >> sure I >> > > >> >> understand, >> > > >> >> > are you suggesting that control messages would be indica= ted >> > by >> > > a >> > > >> >> standard >> > > >> >> > reserved header? By standard, I mean that ALL Kafka >> > consumers >> > > would >> > > >> >> know >> > > >> >> > to handle these messages differently (possibly just >> ignoring >> > > them). >> > > >> >> This >> > > >> >> > would need to be added to the specification of the consu= mer >> > > protocol >> > > >> so >> > > >> >> > that all Kafka clients implement it, right? I think it'= s a >> > > good idea >> > > >> >> but >> > > >> >> > just checking. >> > > >> >> > >> > > >> >> > Cheers, >> > > >> >> > >> > > >> >> > Roger >> > > >> >> > >> > > >> >> > >> > > >> >> > On Wed, Dec 14, 2016 at 9:51 AM, Matthias J. Sax < >> > > >> matthias@confluent.io >> > > >> >> > >> > > >> >> > wrote: >> > > >> >> > >> > > >> >> >> Yes and no. I did overload the term "control message". >> > > >> >> >> >> > > >> >> >> EOS control messages are for client-broker communicatio= n >> and >> > > thus >> > > >> never >> > > >> >> >> exposed to any application. And I think this is a good >> > design >> > > because >> > > >> >> >> broker needs to understand those control messages. Thus= , >> > this >> > > should >> > > >> be >> > > >> >> >> a protocol change. >> > > >> >> >> >> > > >> >> >> The type of control messages I have in mind are for >> > > client-client >> > > >> >> >> (application-application) communication and the broker = is >> > > agnostic to >> > > >> >> >> them. Thus, it should not be a protocol change. >> > > >> >> >> >> > > >> >> >> >> > > >> >> >> -Matthias >> > > >> >> >> >> > > >> >> >> >> > > >> >> >> >> > > >> >> >> On 12/14/16 9:42 AM, radai wrote: >> > > >> >> >>> arent control messages getting pushed as their own top >> > level >> > > >> protocol >> > > >> >> >>> change (and a fairly massive one) for the transactions >> KIP >> > ? >> > > >> >> >>> >> > > >> >> >>> On Tue, Dec 13, 2016 at 5:54 PM, Matthias J. Sax < >> > > >> >> matthias@confluent.io> >> > > >> >> >>> wrote: >> > > >> >> >>> >> > > >> >> >>>> Hi, >> > > >> >> >>>> >> > > >> >> >>>> I want to add a completely new angle to this discussi= on. >> > > For this, >> > > >> I >> > > >> >> >>>> want to propose an extension for the headers feature >> that >> > > enables >> > > >> new >> > > >> >> >>>> uses cases -- and those new use cases might convince >> > people >> > > to >> > > >> >> support >> > > >> >> >>>> headers (of course including the larger scoped >> proposal). >> > > >> >> >>>> >> > > >> >> >>>> Extended Proposal: >> > > >> >> >>>> >> > > >> >> >>>> Allow messages with a certain header key to be specia= l >> > > "control >> > > >> >> >>>> messages" (w/ o w/o payload) that are not exposed to = an >> > > application >> > > >> >> via >> > > >> >> >>>> .poll(). >> > > >> >> >>>> >> > > >> >> >>>> Thus, a consumer client would automatically skip over >> > those >> > > >> >> messages. If >> > > >> >> >>>> an application knows about embedded control messages,= it >> > > can "sing >> > > >> >> up" >> > > >> >> >>>> to those messages by the consumer client and either g= et >> a >> > > callback >> > > >> or >> > > >> >> >>>> the consumer auto-drop for this messages gets disable= d >> > > (allowing to >> > > >> >> >>>> consumer those messages via poll()). >> > > >> >> >>>> >> > > >> >> >>>> (The details need further considerations/discussion. = I >> > just >> > > want to >> > > >> >> >>>> sketch the main idea.) >> > > >> >> >>>> >> > > >> >> >>>> Usage: >> > > >> >> >>>> >> > > >> >> >>>> There is a shared topic (ie, used by multiple >> > applications) >> > > and a >> > > >> >> >>>> producer application wants to embed a special message= in >> > > the topic >> > > >> >> for a >> > > >> >> >>>> dedicated consumer application. Because only one >> > > application will >> > > >> >> >>>> understand this message, it cannot be a regular messa= ge >> as >> > > this >> > > >> would >> > > >> >> >>>> break all applications that do not understand this >> > message. >> > > The >> > > >> >> producer >> > > >> >> >>>> application would set a special metadata key and no >> > consumer >> > > >> >> application >> > > >> >> >>>> would see this control message by default because the= y >> did >> > > not >> > > >> enable >> > > >> >> >>>> their consumer client to return this message in poll(= ) >> > (and >> > > the >> > > >> >> client >> > > >> >> >>>> would just drop this message with special metadata ke= y). >> > > Only the >> > > >> >> single >> > > >> >> >>>> application that should receive this message, will >> > > subscribe to >> > > >> this >> > > >> >> >>>> message on its consumer client and process it. >> > > >> >> >>>> >> > > >> >> >>>> >> > > >> >> >>>> Concrete Use Case: Kafka Streams >> > > >> >> >>>> >> > > >> >> >>>> In Kafka Streams, we would like to propagate "control >> > > messages" >> > > >> from >> > > >> >> >>>> subtopology to subtopology. There are multiple scenar= ios >> > > for which >> > > >> >> this >> > > >> >> >>>> would be useful. For example, currently we do not >> > guarantee >> > > a >> > > >> >> >>>> "consistent shutdown" of an application. By this, I m= ean >> > > that input >> > > >> >> >>>> records might not be completely processed by the whol= e >> > > topology >> > > >> >> because >> > > >> >> >>>> the application shutdown happens "in between" and an >> > > intermediate >> > > >> >> result >> > > >> >> >>>> topic gets "stock" in an intermediate topic. Thus, a >> user >> > > would see >> > > >> >> an >> > > >> >> >>>> committed offset of the source topic of the applicati= on, >> > > but no >> > > >> >> >>>> corresponding result record in the output topic. >> > > >> >> >>>> >> > > >> >> >>>> Having "shutdown markers" would allow us, to first st= op >> > the >> > > >> upstream >> > > >> >> >>>> subtopology and write this marker into the intermedia= te >> > > topic and >> > > >> the >> > > >> >> >>>> downstream subtopology would only shut down itself af= ter >> > is >> > > sees >> > > >> the >> > > >> >> >>>> "shutdown marker". Thus, we can guarantee on shutdown= , >> > that >> > > no >> > > >> >> >>>> "in-flight" messages got stuck in intermediate topics= . >> > > >> >> >>>> >> > > >> >> >>>> >> > > >> >> >>>> A similar usage would be for KIP-95 (Incremental Batc= h >> > > Processing). >> > > >> >> >>>> There was a discussion about the proposed metadata >> topic, >> > > and we >> > > >> >> could >> > > >> >> >>>> avoid this metadata topic if we would have "control >> > > messages". >> > > >> >> >>>> >> > > >> >> >>>> >> > > >> >> >>>> Right now, we cannot insert an "application control >> > message" >> > > >> because >> > > >> >> >>>> Kafka Streams does not own all topics it read/writes = and >> > > thus might >> > > >> >> >>>> break other consumer application (as described above)= if >> > we >> > > inject >> > > >> >> >>>> random messages that are not understood by other apps= . >> > > >> >> >>>> >> > > >> >> >>>> >> > > >> >> >>>> Of course, one can work around "embedded control >> messaged" >> > > by using >> > > >> >> an >> > > >> >> >>>> additional topic to propagate control messaged betwee= n >> > > application >> > > >> >> (as >> > > >> >> >>>> suggestion in KIP-95 via a metadata topic for Kafka >> > > Streams). But >> > > >> >> there >> > > >> >> >>>> are major concerns about adding this metadata topic i= n >> the >> > > KIP and >> > > >> >> this >> > > >> >> >>>> shows that other application that need a similar patt= ern >> > > might >> > > >> profit >> > > >> >> >>>> from topic embedded "control messages", too. >> > > >> >> >>>> >> > > >> >> >>>> >> > > >> >> >>>> One last important consideration: those "control >> messages" >> > > are used >> > > >> >> for >> > > >> >> >>>> client to client communication and are not understood= by >> > the >> > > >> broker. >> > > >> >> >>>> Thus, those messages should not be enabled within the >> > > message >> > > >> format >> > > >> >> >>>> (c.f. tombstone flag -- KIP-87). However, "client lan= d" >> > > record >> > > >> >> headers >> > > >> >> >>>> would be a nice way to implement them. Because KIP-82 >> did >> > > consider >> > > >> >> key >> > > >> >> >>>> namespaces for metatdata keys, this extension should = not >> > be >> > > an own >> > > >> >> KIP >> > > >> >> >>>> but should be included in KIP-82 to reserve a namespa= ce >> > for >> > > >> "control >> > > >> >> >>>> message" in the first place. >> > > >> >> >>>> >> > > >> >> >>>> >> > > >> >> >>>> Sorry for the long email... Looking forward to your >> > > feedback. >> > > >> >> >>>> >> > > >> >> >>>> >> > > >> >> >>>> -Matthias >> > > >> >> >>>> >> > > >> >> >>>> >> > > >> >> >>>> >> > > >> >> >>>> >> > > >> >> >>>> >> > > >> >> >>>> >> > > >> >> >>>> >> > > >> >> >>>> >> > > >> >> >>>> >> > > >> >> >>>> On 12/8/16 12:12 AM, Michael Pearce wrote: >> > > >> >> >>>>> Hi Jun >> > > >> >> >>>>> >> > > >> >> >>>>> 100) each time a transaction exits a jvm for a remot= e >> > > system >> > > >> >> (HTTP/JMS/ >> > > >> >> >>>> Hopefully one day kafka) the APM tools stich in a uni= que >> > id >> > > >> (though I >> > > >> >> >>>> believe it contains the end2end uuid embedded in this >> id), >> > > on >> > > >> >> receiving >> > > >> >> >> the >> > > >> >> >>>> message at the receiving JVM the apm code takes this >> out, >> > > and >> > > >> >> continues >> > > >> >> >> its >> > > >> >> >>>> tracing on the that new thread. Both JVM=E2=80=99s (a= nd other >> > > languages the >> > > >> >> APM >> > > >> >> >>>> tool supports) send this data async back to the centr= al >> > > controllers >> > > >> >> >> where >> > > >> >> >>>> the stiching togeather occurs. For this they need som= e >> > > header space >> > > >> >> for >> > > >> >> >>>> them to put this id. >> > > >> >> >>>>> >> > > >> >> >>>>> 101) Yes indeed we have a business transaction Id in >> the >> > > payload. >> > > >> >> >> Though >> > > >> >> >>>> this is a system level tracing, that we need to have >> marry >> > > up. Also >> > > >> >> as >> > > >> >> >> per >> > > >> >> >>>> note on end2end encryption we=E2=80=99d be unable to = prove the >> > flow >> > > if the >> > > >> >> >> payload >> > > >> >> >>>> is encrypted as we=E2=80=99d not have access to this = at certain >> > > points of >> > > >> the >> > > >> >> >> flow >> > > >> >> >>>> through the infrastructure/platform. >> > > >> >> >>>>> >> > > >> >> >>>>> >> > > >> >> >>>>> 103) As said we use this mechanism in IG very >> > > successfully, as >> > > >> >> stated >> > > >> >> >>>> per key we guarantee the transaction producing app to >> > > handle the >> > > >> >> >>>> transaction of a key at one DC unless at point of >> critical >> > > failure >> > > >> >> >> where we >> > > >> >> >>>> have to flip processing to another. We care about key >> > > ordering. >> > > >> >> >>>>> I disagree on the offset comment for the partition >> > > solution unless >> > > >> >> you >> > > >> >> >>>> do full ISR, or expensive full XA transactions even w= ith >> > > partitions >> > > >> >> you >> > > >> >> >>>> cannot fully guarantee offsets would match. >> > > >> >> >>>>> >> > > >> >> >>>>> 105) Very much so, I need to have access at the >> platform >> > > level to >> > > >> >> the >> > > >> >> >>>> other meta data all mentioned, without having to need= to >> > > have >> > > >> access >> > > >> >> to >> > > >> >> >> the >> > > >> >> >>>> encryption keys of the payload. >> > > >> >> >>>>> >> > > >> >> >>>>> 106) >> > > >> >> >>>>> Techincally yes for AZ/Region/Cluster, but then we= =E2=80=99d >> need >> > > to have >> > > >> a >> > > >> >> >>>> global producerId register which would be very hard t= o >> > > >> >> enforce/ensure is >> > > >> >> >>>> current and correct, just to understand the message >> > origins >> > > of its >> > > >> >> >>>> region/az/cluster for routing. >> > > >> >> >>>>> The client wrapper version, producerId can be the sa= me, >> > as >> > > >> obviously >> > > >> >> >> the >> > > >> >> >>>> producer could upgrade its wrapper, as such we need t= o >> > know >> > > what >> > > >> >> wrapper >> > > >> >> >>>> version the message is created with. >> > > >> >> >>>>> Likewise the IP address, as stated we can have our >> > > producer move, >> > > >> >> where >> > > >> >> >>>> its IP would change. >> > > >> >> >>>>> >> > > >> >> >>>>> 107) >> > > >> >> >>>>> UUID is set on the message by interceptors before >> actual >> > > producer >> > > >> >> >>>> transport send. This is for platform level message >> dedupe >> > > >> guarantee, >> > > >> >> the >> > > >> >> >>>> business payload should be agnostic to this. Please s= ee >> > > >> >> >>>> https://activemq.apache.org/ >> artemis/docs/1.5.0/duplicate- >> > > det >> > > >> >> ection.html >> > > >> >> >>>> note this is not touching business payloads. >> > > >> >> >>>>> >> > > >> >> >>>>> >> > > >> >> >>>>> >> > > >> >> >>>>> On 06/12/2016, 18:22, "Jun Rao" >> > wrote: >> > > >> >> >>>>> >> > > >> >> >>>>> Hi, Michael, >> > > >> >> >>>>> >> > > >> >> >>>>> Thanks for the reply. I find it very helpful. >> > > >> >> >>>>> >> > > >> >> >>>>> Data lineage: >> > > >> >> >>>>> 100. I'd like to understand the APM use case a b= it >> > > more. It >> > > >> >> sounds >> > > >> >> >>>> like >> > > >> >> >>>>> that those APM plugins can generate a transactio= n >> id >> > > that we >> > > >> >> could >> > > >> >> >>>>> potentially put in the header of every message. = How >> > > would you >> > > >> >> >>>> typically >> > > >> >> >>>>> make use of such transaction ids? Are there othe= r >> > > metadata >> > > >> >> >>>> associated with >> > > >> >> >>>>> the transaction id and if so, how are they >> propagated >> > > >> >> downstream? >> > > >> >> >>>>> >> > > >> >> >>>>> 101. For the finance use case, if the concept of >> > > transaction >> > > >> is >> > > >> >> >>>> important, >> > > >> >> >>>>> wouldn't it be typically included in the message >> > > payload >> > > >> >> instead of >> > > >> >> >>>> as an >> > > >> >> >>>>> optional header field? >> > > >> >> >>>>> >> > > >> >> >>>>> 102. The data lineage that Altas and Navigator >> > support >> > > seems >> > > >> to >> > > >> >> be >> > > >> >> >>>> at the >> > > >> >> >>>>> dataset level, not per record level? So, not sur= e >> if >> > > per >> > > >> message >> > > >> >> >>>> headers >> > > >> >> >>>>> are relevant there. >> > > >> >> >>>>> >> > > >> >> >>>>> Mirroring: >> > > >> >> >>>>> 103. The benefit of using separate partitions is >> that >> > > it >> > > >> >> >> potentially >> > > >> >> >>>> makes >> > > >> >> >>>>> it easy to preserve offsets during mirroring. Th= is >> > > will make >> > > >> it >> > > >> >> >>>> easier for >> > > >> >> >>>>> consumer to switch clusters. Currently, the >> consumers >> > > can >> > > >> switch >> > > >> >> >>>> clusters >> > > >> >> >>>>> by using the timestampToOffset() api, but it has= to >> > > deal with >> > > >> >> >>>> duplicates. >> > > >> >> >>>>> Good point on the issue with log compact and I a= m >> not >> > > sure how >> > > >> >> to >> > > >> >> >>>> address >> > > >> >> >>>>> this. However, even if we mirror into the existi= ng >> > > partitions, >> > > >> >> the >> > > >> >> >>>> ordering >> > > >> >> >>>>> for messages generated from different clusters >> seems >> > > >> >> >>>> non-deterministic >> > > >> >> >>>>> anyway. So, it seems that the consumers already >> have >> > > to deal >> > > >> >> with >> > > >> >> >>>> that? If >> > > >> >> >>>>> a topic is compacted, does that mean which messa= ges >> > are >> > > >> >> preserved >> > > >> >> >> is >> > > >> >> >>>> also >> > > >> >> >>>>> non-deterministic across clusters? >> > > >> >> >>>>> >> > > >> >> >>>>> 104. Good point on partition key. >> > > >> >> >>>>> >> > > >> >> >>>>> End-to-end encryption: >> > > >> >> >>>>> 105. So, it seems end-to-end encryption is usefu= l. >> > Are >> > > headers >> > > >> >> >>>> useful there? >> > > >> >> >>>>> >> > > >> >> >>>>> Auditing: >> > > >> >> >>>>> 106. It seems other than the UUID, all other >> metadata >> > > are per >> > > >> >> >>>> producer? >> > > >> >> >>>>> >> > > >> >> >>>>> EOS: >> > > >> >> >>>>> 107. How are those UUIDs generated? I am not sur= e >> if >> > > they can >> > > >> be >> > > >> >> >>>> generated >> > > >> >> >>>>> in the producer library. An application may send >> > > messages >> > > >> >> through a >> > > >> >> >>>> load >> > > >> >> >>>>> balancer and on retry, the same message could be >> > > routed to a >> > > >> >> >>>> different >> > > >> >> >>>>> producer instance. So, it seems that the >> application >> > > has to >> > > >> >> >> generate >> > > >> >> >>>> the >> > > >> >> >>>>> UUIDs. In that case, shouldn't the application j= ust >> > > put the >> > > >> >> UUID in >> > > >> >> >>>> the >> > > >> >> >>>>> payload? >> > > >> >> >>>>> >> > > >> >> >>>>> Thanks, >> > > >> >> >>>>> >> > > >> >> >>>>> Jun >> > > >> >> >>>>> >> > > >> >> >>>>> >> > > >> >> >>>>> On Fri, Dec 2, 2016 at 4:57 PM, Michael Pearce < >> > > >> >> >>>> Michael.Pearce@ig.com> >> > > >> >> >>>>> wrote: >> > > >> >> >>>>> >> > > >> >> >>>>> > Hi Jun. >> > > >> >> >>>>> > >> > > >> >> >>>>> > Per Transaction Tracing / Data Lineage. >> > > >> >> >>>>> > >> > > >> >> >>>>> > As Stated in the KIP this has the first use ca= se >> of >> > > how many >> > > >> >> APM >> > > >> >> >>>> tools now >> > > >> >> >>>>> > work. >> > > >> >> >>>>> > I would find it impossible for any one to argu= e >> > this >> > > is not >> > > >> >> >>>> important or a >> > > >> >> >>>>> > niche market as it has its own gartner report = for >> > > this >> > > >> space. >> > > >> >> >> Such >> > > >> >> >>>>> > companies as Appdynamics, NewRelic, Dynatrace, >> > > Hawqular are >> > > >> >> but a >> > > >> >> >>>> few. >> > > >> >> >>>>> > >> > > >> >> >>>>> > Likewise these APM tools can help very rapidly >> > track >> > > down >> > > >> >> issues >> > > >> >> >>>> and >> > > >> >> >>>>> > automatically capture metrics, perform actions >> > based >> > > on >> > > >> >> >> unexpected >> > > >> >> >>>> behavior >> > > >> >> >>>>> > to auto recover services. >> > > >> >> >>>>> > >> > > >> >> >>>>> > Before mentioning looking at aggregated stats,= in >> > > these >> > > >> cases >> > > >> >> >> where >> > > >> >> >>>>> > actually on critical flows we cannot afford to >> have >> > > >> aggregated >> > > >> >> >>>> rolled up >> > > >> >> >>>>> > stats only. >> > > >> >> >>>>> > >> > > >> >> >>>>> > With the APM tool we use its actually able to >> > detect >> > > a >> > > >> single >> > > >> >> >>>> transaction >> > > >> >> >>>>> > failure and capture the thread traces in the J= VM >> > > where it >> > > >> >> failed >> > > >> >> >>>> and >> > > >> >> >>>>> > everything for us, to the point it sends us >> alerts >> > > where we >> > > >> >> have >> > > >> >> >>>> this >> > > >> >> >>>>> > giving the line number of the code that caused >> it, >> > > the >> > > >> >> >> transaction >> > > >> >> >>>> trace >> > > >> >> >>>>> > through all the services and endpoints >> (supported) >> > > upto the >> > > >> >> point >> > > >> >> >>>> of >> > > >> >> >>>>> > failure, it can also capture the data in and o= ut >> > (so >> > > we can >> > > >> >> >>>> replay). >> > > >> >> >>>>> > Because atm Kafka doesn=E2=80=99t support us b= eing able >> to >> > > stich in >> > > >> >> these >> > > >> >> >>>> tracing >> > > >> >> >>>>> > transaction ids natively, we cannot get these >> > > benefits as >> > > >> >> such is >> > > >> >> >>>> limiting >> > > >> >> >>>>> > our ability support apps and monitor them to t= he >> > same >> > > >> >> standards >> > > >> >> >> we >> > > >> >> >>>> come to >> > > >> >> >>>>> > expect when on a kafka flow. >> > > >> >> >>>>> > >> > > >> >> >>>>> > This actually ties in with Data Lineage, as th= e >> > same >> > > tracing >> > > >> >> can >> > > >> >> >>>> be used >> > > >> >> >>>>> > to back stich this. Essentially many times due= to >> > > the sums >> > > >> of >> > > >> >> >> money >> > > >> >> >>>>> > involved there are disputes, and typically as = a >> > > financial >> > > >> >> >>>> institute the >> > > >> >> >>>>> > easiest and cleanest way to prove when dispute= s >> > > arise is to >> > > >> >> >>>> present the >> > > >> >> >>>>> > actual flow and processes involved in a >> > transaction. >> > > >> >> >>>>> > >> > > >> >> >>>>> > Likewise as Hadoop matures its evident this ca= se >> is >> > > >> >> important, as >> > > >> >> >>>> tools >> > > >> >> >>>>> > such as Atlas (Hortonworks led) and Navigator >> > > (cloudera led) >> > > >> >> are >> > > >> >> >>>> evident >> > > >> >> >>>>> > also I believe the importance here is very muc= h >> NOT >> > > just a >> > > >> >> >>>> financial issue. >> > > >> >> >>>>> > >> > > >> >> >>>>> > From a MDM point of view any company wanting t= o >> > care >> > > about >> > > >> >> Data >> > > >> >> >>>> Quality >> > > >> >> >>>>> > and Data Governance - Data Lineage is a key pi= ece >> > in >> > > this >> > > >> >> puzzle. >> > > >> >> >>>>> > >> > > >> >> >>>>> > >> > > >> >> >>>>> > >> > > >> >> >>>>> > RE Mirroring, >> > > >> >> >>>>> > >> > > >> >> >>>>> > As per the KIP in-fact this is exactly what we= do >> > re >> > > cluster >> > > >> >> id, >> > > >> >> >>>> to mirror >> > > >> >> >>>>> > a network of clusters between AZ=E2=80=99s / R= egions. We >> > > know a >> > > >> >> >>>> transaction for a >> > > >> >> >>>>> > key will be done within a AZ/Region, as such = we >> > > know the >> > > >> >> write >> > > >> >> >> to >> > > >> >> >>>> kafka >> > > >> >> >>>>> > would be ordered per key. But we need eventual >> view >> > > of that >> > > >> >> >> across >> > > >> >> >>>> in our >> > > >> >> >>>>> > other regions/az=E2=80=99s. When we have compl= ete AZ or >> > > Region >> > > >> >> failure we >> > > >> >> >>>> know >> > > >> >> >>>>> > there will be a brief interruption whilst thos= e >> > > transactions >> > > >> >> are >> > > >> >> >>>> moved to >> > > >> >> >>>>> > another region but we expect after it to >> continue. >> > > >> >> >>>>> > >> > > >> >> >>>>> > As mentioned having separate Partions to do th= is >> > > starts to >> > > >> get >> > > >> >> >>>>> > ugly/complicated for us: >> > > >> >> >>>>> > how would I do compaction where a key is in tw= o >> > > partitions? >> > > >> >> >>>>> > How do we balance consumers so where multiple >> > > partitions >> > > >> with >> > > >> >> the >> > > >> >> >>>> same key >> > > >> >> >>>>> > goto the same consumer >> > > >> >> >>>>> > What do you do if cluster 1 has 5 partitions b= ut >> > > cluster 20 >> > > >> >> has >> > > >> >> >> 10 >> > > >> >> >>>> because >> > > >> >> >>>>> > its larger kit in our more core DC=E2=80=99s, = as such key >> > to >> > > >> partition >> > > >> >> >>>> mappings for >> > > >> >> >>>>> > consumers get even more complicated. >> > > >> >> >>>>> > What do you do if we add or remove a complete >> > region >> > > >> >> >>>>> > >> > > >> >> >>>>> > Where as simple mirror will work we just need = to >> > > ensure we >> > > >> >> don=E2=80=99t >> > > >> >> >>>> have a >> > > >> >> >>>>> > cycle which we can do with clusterId. >> > > >> >> >>>>> > >> > > >> >> >>>>> > We even have started to look at shortest path >> > mirror >> > > routing >> > > >> >> >> based >> > > >> >> >>>> on >> > > >> >> >>>>> > clusterId, if we also had the region and az in= fo >> on >> > > the >> > > >> >> >> originating >> > > >> >> >>>>> > message, this we have not implemented but some >> > ideas >> > > come >> > > >> from >> > > >> >> >>>> network >> > > >> >> >>>>> > routing, and also the dispatcher router in apa= che >> > > qpid. >> > > >> >> >>>>> > >> > > >> >> >>>>> > Also we need to have data perimeters e.g. cert= ain >> > > data >> > > >> cannot >> > > >> >> >> leave >> > > >> >> >>>>> > certain countries borders. We want this all >> > > automated so >> > > >> that >> > > >> >> at >> > > >> >> >>>> the >> > > >> >> >>>>> > platform level without having to touch or look= at >> > the >> > > >> business >> > > >> >> >>>> data inside >> > > >> >> >>>>> > we can have headers we can put tags into so th= at >> we >> > > can >> > > >> ensure >> > > >> >> >>>> this doesn=E2=80=99t >> > > >> >> >>>>> > occur when we mirror. (actually links in to da= ta >> > > lineage / >> > > >> >> >> tracing >> > > >> >> >>>> as again >> > > >> >> >>>>> > we need to tag messages at a platform level) >> > > Examples are we >> > > >> >> are >> > > >> >> >>>> not >> > > >> >> >>>>> > allowed Private customer details to leave >> > > Switzerland, yet >> > > >> we >> > > >> >> >> need >> > > >> >> >>>> those >> > > >> >> >>>>> > systems integrated. >> > > >> >> >>>>> > >> > > >> >> >>>>> > Lastly around mirroring we have a partionKey >> field, >> > > as the >> > > >> key >> > > >> >> >>>> used for >> > > >> >> >>>>> > portioning logic !=3D compaction key all the t= ime >> but >> > > we want >> > > >> to >> > > >> >> >>>> preserve it >> > > >> >> >>>>> > for when we mirror so that if source cluster >> > > partition count >> > > >> >> !=3D >> > > >> >> >>>> destination >> > > >> >> >>>>> > cluster partition count we can honour the same >> > > partitioning >> > > >> >> >> logic. >> > > >> >> >>>>> > >> > > >> >> >>>>> > >> > > >> >> >>>>> > >> > > >> >> >>>>> > RE End 2 End encryption >> > > >> >> >>>>> > >> > > >> >> >>>>> > As I believe mentioned just before, the soluti= on >> > you >> > > mention >> > > >> >> just >> > > >> >> >>>> doesn=E2=80=99t >> > > >> >> >>>>> > cut the mustard these days with many regulator= s. >> An >> > > >> operations >> > > >> >> >>>> person with >> > > >> >> >>>>> > access to the box should not be able to have >> access >> > > to the >> > > >> >> data. >> > > >> >> >>>> Many now >> > > >> >> >>>>> > actually impose quite literally the >> implementation >> > > expected >> > > >> >> being >> > > >> >> >>>> end2end >> > > >> >> >>>>> > encryption for certain data (Singapore for us = is >> > one >> > > that I >> > > >> am >> > > >> >> >>>> most aware >> > > >> >> >>>>> > of). In fact we=E2=80=99re even now needing en= crypt the >> > data >> > > and >> > > >> store >> > > >> >> >> the >> > > >> >> >>>> keys in >> > > >> >> >>>>> > HSM modules. >> > > >> >> >>>>> > >> > > >> >> >>>>> > Likewise the performance penalty on encrypting >> > > decrypting as >> > > >> >> you >> > > >> >> >>>> produce >> > > >> >> >>>>> > over wire, then again encrypt decrypt as the d= ata >> > is >> > > stored >> > > >> on >> > > >> >> >> the >> > > >> >> >>>> brokers >> > > >> >> >>>>> > disks and back again, then again encrypted and >> > > decrypted >> > > >> back >> > > >> >> >> over >> > > >> >> >>>> the wire >> > > >> >> >>>>> > each time for each consumer all adds up, ignor= ing >> > > this >> > > >> >> doubling >> > > >> >> >>>> with mirror >> > > >> >> >>>>> > makers etc. simply encrypting the value once o= n >> > > write by the >> > > >> >> >>>> client and >> > > >> >> >>>>> > again decrypting on consume by the consumer is >> far >> > > more >> > > >> >> >>>> performant, but >> > > >> >> >>>>> > then the routing and platform meta data needs = to >> be >> > > separate >> > > >> >> >> (thus >> > > >> >> >>>> headers) >> > > >> >> >>>>> > >> > > >> >> >>>>> > >> > > >> >> >>>>> > >> > > >> >> >>>>> > RE Auditing: >> > > >> >> >>>>> > >> > > >> >> >>>>> > Our Auditing needs are: >> > > >> >> >>>>> > Producer Id, >> > > >> >> >>>>> > Origin Cluster Id that message first produced >> into >> > > >> >> >>>>> > Origin AZ =E2=80=93 agreed we can derive this = if we have >> > > cluster id, >> > > >> >> but >> > > >> >> >>>> it makes >> > > >> >> >>>>> > resolving this for audit reporting a lot easie= r. >> > > >> >> >>>>> > Origin Region =E2=80=93 agreed we can derive t= his if we >> > have >> > > cluster >> > > >> >> id, >> > > >> >> >>>> but it >> > > >> >> >>>>> > makes resolving this for audit reporting a lot >> > > easier. >> > > >> >> >>>>> > Unique Message Identification (this is not the >> same >> > > as >> > > >> >> >> transaction >> > > >> >> >>>>> > tracing) =E2=80=93 note offset and partition a= re not the >> > > same, as >> > > >> >> when we >> > > >> >> >>>> mirror or >> > > >> >> >>>>> > have for what ever system failure duplicate se= nd, >> > > >> >> >>>>> > Custom Client wrapper version (where >> organizations >> > > have to >> > > >> >> wrap >> > > >> >> >>>> the kafka >> > > >> >> >>>>> > client for added features) so we know what >> version >> > > of the >> > > >> >> wrapper >> > > >> >> >>>> is used >> > > >> >> >>>>> > Producer IP address (in case of clients being = in >> > our >> > > vm/open >> > > >> >> >> stack >> > > >> >> >>>> infra >> > > >> >> >>>>> > where they can move around, producer id will s= tay >> > > the same >> > > >> but >> > > >> >> >>>> this would >> > > >> >> >>>>> > change) >> > > >> >> >>>>> > >> > > >> >> >>>>> > >> > > >> >> >>>>> > >> > > >> >> >>>>> > RE Once and only once delivery case >> > > >> >> >>>>> > >> > > >> >> >>>>> > Using the same Message UUID for auditing we ca= n >> > > achieve this >> > > >> >> >> quite >> > > >> >> >>>> simply. >> > > >> >> >>>>> > >> > > >> >> >>>>> > As per how some other brokers do this (cough >> qpid, >> > > artemis) >> > > >> >> >>>> message uuid >> > > >> >> >>>>> > are used to dedupe where message is sent and >> > > produced but >> > > >> the >> > > >> >> >>>> client didn=E2=80=99t >> > > >> >> >>>>> > receive the ack, and there for replays the sen= d, >> by >> > > having a >> > > >> >> >>>> unique message >> > > >> >> >>>>> > id per message, this can be filtered out, on >> > > consumers where >> > > >> >> >>>> message >> > > >> >> >>>>> > delivery may occur twice for what ever reasons= a >> > > message >> > > >> uuid >> > > >> >> can >> > > >> >> >>>> be used >> > > >> >> >>>>> > to remove duplicates being deliverd , like wis= e >> we >> > > can do >> > > >> >> this in >> > > >> >> >>>> the >> > > >> >> >>>>> > mirrormakers so if we detect a dupe message we >> can >> > > avoid >> > > >> >> >>>> replicating it. >> > > >> >> >>>>> > >> > > >> >> >>>>> > >> > > >> >> >>>>> > >> > > >> >> >>>>> > >> > > >> >> >>>>> > Cheers >> > > >> >> >>>>> > Mike >> > > >> >> >>>>> > >> > > >> >> >>>>> > >> > > >> >> >>>>> > >> > > >> >> >>>>> > On 02/12/2016, 22:09, "Jun Rao" < >> jun@confluent.io> >> > > wrote: >> > > >> >> >>>>> > >> > > >> >> >>>>> > Since this KIP affects message format, wir= e >> > > protocol, >> > > >> >> apis, I >> > > >> >> >>>> think >> > > >> >> >>>>> > it's >> > > >> >> >>>>> > worth spending a bit more time to nail dow= n >> the >> > > concrete >> > > >> >> use >> > > >> >> >>>> cases. It >> > > >> >> >>>>> > would be bad if we add this feature, but w= hen >> > > start >> > > >> >> >>>> implementing it >> > > >> >> >>>>> > for say >> > > >> >> >>>>> > mirroring, we then realize that header is = not >> > > the best >> > > >> >> >>>> approach. >> > > >> >> >>>>> > Initially, >> > > >> >> >>>>> > I thought I was convinced of the use cases= of >> > > headers >> > > >> and >> > > >> >> was >> > > >> >> >>>> trying to >> > > >> >> >>>>> > write down a few use cases to convince >> others. >> > > That's >> > > >> >> when I >> > > >> >> >>>> became >> > > >> >> >>>>> > less >> > > >> >> >>>>> > certain. For me to be convinced, I just wa= nt >> to >> > > see two >> > > >> >> >> strong >> > > >> >> >>>> use >> > > >> >> >>>>> > cases >> > > >> >> >>>>> > (instead of 10 maybe use cases) in the >> > > third-party >> > > >> space. >> > > >> >> The >> > > >> >> >>>> reason is >> > > >> >> >>>>> > that when we discussed the use cases withi= n a >> > > company, >> > > >> >> often >> > > >> >> >>>> it ends >> > > >> >> >>>>> > with >> > > >> >> >>>>> > "we can't force everyone to use this stand= ard >> > > since we >> > > >> may >> > > >> >> >>>> have to >> > > >> >> >>>>> > integrate with third-party tools". >> > > >> >> >>>>> > >> > > >> >> >>>>> > At present, I am not sure why headers are >> > useful >> > > for >> > > >> >> things >> > > >> >> >>>> like >> > > >> >> >>>>> > schemaId >> > > >> >> >>>>> > or encryption. In order to do anything use= ful >> > to >> > > the >> > > >> >> value, >> > > >> >> >>>> one needs >> > > >> >> >>>>> > to >> > > >> >> >>>>> > know the schemaId or how data is encrypted= , >> but >> > > header >> > > >> is >> > > >> >> >>>> optional. >> > > >> >> >>>>> > But, I >> > > >> >> >>>>> > can be convinced if someone (Radai, Sean, >> > Todd?) >> > > >> provides >> > > >> >> >> more >> > > >> >> >>>> details >> > > >> >> >>>>> > on >> > > >> >> >>>>> > the argument. >> > > >> >> >>>>> > >> > > >> >> >>>>> > I am not very sure header is the best >> approach >> > > for >> > > >> >> mirroring >> > > >> >> >>>> either. If >> > > >> >> >>>>> > someone has thought about this more, I'd b= e >> > > happy to >> > > >> hear. >> > > >> >> >>>>> > >> > > >> >> >>>>> > I can see the data lineage use case. I am >> just >> > > not sure >> > > >> >> how >> > > >> >> >>>> widely >> > > >> >> >>>>> > applicable this is. If someone familiar wi= th >> > > this space >> > > >> >> can >> > > >> >> >>>> justify >> > > >> >> >>>>> > this is >> > > >> >> >>>>> > a significant use case, say in the finance >> > > industry, >> > > >> this >> > > >> >> >>>> would be a >> > > >> >> >>>>> > strong >> > > >> >> >>>>> > use case. >> > > >> >> >>>>> > >> > > >> >> >>>>> > I can see the auditing use case. I am just >> not >> > > sure if a >> > > >> >> >> native >> > > >> >> >>>>> > producer id >> > > >> >> >>>>> > solves that problem. If there are addition= al >> > > metadata >> > > >> >> that's >> > > >> >> >>>> worth >> > > >> >> >>>>> > collecting but not covered by the producer >> id, >> > > that >> > > >> would >> > > >> >> >> make >> > > >> >> >>>> this a >> > > >> >> >>>>> > strong use case. >> > > >> >> >>>>> > >> > > >> >> >>>>> > Thanks, >> > > >> >> >>>>> > >> > > >> >> >>>>> > Jun >> > > >> >> >>>>> > >> > > >> >> >>>>> > >> > > >> >> >>>>> > On Fri, Dec 2, 2016 at 1:41 PM, radai < >> > > >> >> >>>> radai.rosenblatt@gmail.com> >> > > >> >> >>>>> > wrote: >> > > >> >> >>>>> > >> > > >> >> >>>>> > > this KIP is about enabling headers, noth= ing >> > > more >> > > >> nothing >> > > >> >> >>>> less - so >> > > >> >> >>>>> > no, >> > > >> >> >>>>> > > broker-side use of headers is not in the >> KIP >> > > scope. >> > > >> >> >>>>> > > >> > > >> >> >>>>> > > obviously though, once you have headers >> > > potential use >> > > >> >> cases >> > > >> >> >>>> could >> > > >> >> >>>>> > include >> > > >> >> >>>>> > > broker-side header-aware interceptors >> (which >> > > would be >> > > >> >> the >> > > >> >> >>>> topic of >> > > >> >> >>>>> > other >> > > >> >> >>>>> > > future KIPs). >> > > >> >> >>>>> > > >> > > >> >> >>>>> > > a trivially clear use case (to me) would= be >> > > using such >> > > >> >> >>>> broker-side >> > > >> >> >>>>> > > interceptors to enforce compliance with >> > > organizational >> > > >> >> >>>> policies - it >> > > >> >> >>>>> > would >> > > >> >> >>>>> > > make our SREs lives much easier if inste= ad >> of >> > > >> >> retroactively >> > > >> >> >>>>> > discovering >> > > >> >> >>>>> > > "rogue" topics/users those messages woul= d >> > have >> > > been >> > > >> >> >> rejected >> > > >> >> >>>>> > up-front. >> > > >> >> >>>>> > > >> > > >> >> >>>>> > > the kafka broker code is lacking any suc= h >> > > >> extensibility >> > > >> >> >>>> support >> > > >> >> >>>>> > (beyond >> > > >> >> >>>>> > > maybe authorizer) which is why these use >> > cases >> > > were >> > > >> left >> > > >> >> >> out >> > > >> >> >>>> of the >> > > >> >> >>>>> > "case >> > > >> >> >>>>> > > for headers" doc - broker extensibility = is >> a >> > > separate >> > > >> >> >>>> discussion. >> > > >> >> >>>>> > > >> > > >> >> >>>>> > > On Fri, Dec 2, 2016 at 12:59 PM, Gwen >> > Shapira < >> > > >> >> >>>> gwen@confluent.io> >> > > >> >> >>>>> > wrote: >> > > >> >> >>>>> > > >> > > >> >> >>>>> > > > Woah, I wasn't aware this is something >> > we'll >> > > do. It >> > > >> >> >> wasn't >> > > >> >> >>>> in the >> > > >> >> >>>>> > KIP, >> > > >> >> >>>>> > > > right? >> > > >> >> >>>>> > > > >> > > >> >> >>>>> > > > I guess we could do it the same way AC= Ls >> > > currently >> > > >> >> work. >> > > >> >> >>>>> > > > I had in mind something that will allo= w >> > > admins to >> > > >> >> apply >> > > >> >> >>>> rules to >> > > >> >> >>>>> > the >> > > >> >> >>>>> > > > new create/delete/config topic APIs. S= o >> > Todd >> > > can >> > > >> >> decide >> > > >> >> >> to >> > > >> >> >>>> reject >> > > >> >> >>>>> > > > "create topic" requests that ask for m= ore >> > > than 40 >> > > >> >> >>>> partitions, or >> > > >> >> >>>>> > > > require exactly 3 replicas, or no more >> than >> > > 50GB >> > > >> >> >> partition >> > > >> >> >>>> size, >> > > >> >> >>>>> > etc. >> > > >> >> >>>>> > > > >> > > >> >> >>>>> > > > ACLs were added a bit ad-hoc, if we ar= e >> > > planning to >> > > >> >> apply >> > > >> >> >>>> more >> > > >> >> >>>>> > rules >> > > >> >> >>>>> > > > to requests (and I think we should), w= e >> may >> > > want a >> > > >> bit >> > > >> >> >>>> more generic >> > > >> >> >>>>> > > > design around that. >> > > >> >> >>>>> > > > >> > > >> >> >>>>> > > > On Fri, Dec 2, 2016 at 7:16 AM, radai = < >> > > >> >> >>>> radai.rosenblatt@gmail.com> >> > > >> >> >>>>> > > wrote: >> > > >> >> >>>>> > > > > "wouldn't you be in the business of >> > making >> > > sure >> > > >> >> >> everyone >> > > >> >> >>>> uses >> > > >> >> >>>>> > them >> > > >> >> >>>>> > > > > properly?" >> > > >> >> >>>>> > > > > >> > > >> >> >>>>> > > > > thats where a broker-side plugin wou= ld >> > > come handy >> > > >> - >> > > >> >> any >> > > >> >> >>>> incoming >> > > >> >> >>>>> > > message >> > > >> >> >>>>> > > > > that does not conform to org policy >> (read >> > > - does >> > > >> not >> > > >> >> >>>> have the >> > > >> >> >>>>> > proper >> > > >> >> >>>>> > > > > headers) gets thrown out (with an er= ror >> > > returned >> > > >> to >> > > >> >> >> user) >> > > >> >> >>>>> > > > > >> > > >> >> >>>>> > > > > On Thu, Dec 1, 2016 at 8:44 PM, Todd >> > > Palino < >> > > >> >> >>>> tpalino@gmail.com> >> > > >> >> >>>>> > wrote: >> > > >> >> >>>>> > > > > >> > > >> >> >>>>> > > > >> Come on, I=E2=80=99ve done at least= 2 talks on >> > > this one >> > > >> :) >> > > >> >> >>>>> > > > >> >> > > >> >> >>>>> > > > >> Producing counts to a topic is part= of >> > > it, but >> > > >> >> that=E2=80=99s >> > > >> >> >>>> only >> > > >> >> >>>>> > part. So >> > > >> >> >>>>> > > you >> > > >> >> >>>>> > > > >> count you have 100 messages in topi= c >> A. >> > > When you >> > > >> >> >> mirror >> > > >> >> >>>> topic A >> > > >> >> >>>>> > to >> > > >> >> >>>>> > > > another >> > > >> >> >>>>> > > > >> cluster, you have 99 messages. Wher= e >> was >> > > your >> > > >> >> problem? >> > > >> >> >>>> Or >> > > >> >> >>>>> > worse, you >> > > >> >> >>>>> > > > have >> > > >> >> >>>>> > > > >> 100 messages, but one producer >> > duplicated >> > > >> messages >> > > >> >> and >> > > >> >> >>>> another >> > > >> >> >>>>> > one >> > > >> >> >>>>> > > lost >> > > >> >> >>>>> > > > >> messages. You need details about wh= ere >> > the >> > > >> message >> > > >> >> >> came >> > > >> >> >>>> from in >> > > >> >> >>>>> > order >> > > >> >> >>>>> > > to >> > > >> >> >>>>> > > > >> pinpoint problems when they happen. >> > Source >> > > >> producer >> > > >> >> >>>> info, where >> > > >> >> >>>>> > it was >> > > >> >> >>>>> > > > >> produced into your infrastructure, = and >> > > when it >> > > >> was >> > > >> >> >>>> produced. >> > > >> >> >>>>> > This >> > > >> >> >>>>> > > > requires >> > > >> >> >>>>> > > > >> you to add the information to the >> > message. >> > > >> >> >>>>> > > > >> >> > > >> >> >>>>> > > > >> And yes, you still need to maintain >> your >> > > clients. >> > > >> >> So >> > > >> >> >>>> maybe my >> > > >> >> >>>>> > original >> > > >> >> >>>>> > > > >> example was not the best. My though= ts >> on >> > > not >> > > >> >> wanting >> > > >> >> >> to >> > > >> >> >>>> be >> > > >> >> >>>>> > responsible >> > > >> >> >>>>> > > > for >> > > >> >> >>>>> > > > >> message formats stands, because tha= t=E2=80=99s >> > > very much >> > > >> >> >>>> separate from >> > > >> >> >>>>> > the >> > > >> >> >>>>> > > > client. >> > > >> >> >>>>> > > > >> As you know, we have our own intern= al >> > > client >> > > >> >> library >> > > >> >> >>>> that can >> > > >> >> >>>>> > insert >> > > >> >> >>>>> > > the >> > > >> >> >>>>> > > > >> right headers, and right now insert= s >> the >> > > right >> > > >> >> audit >> > > >> >> >>>>> > information into >> > > >> >> >>>>> > > > the >> > > >> >> >>>>> > > > >> message fields. If they exist, and >> > > assuming the >> > > >> >> >> message >> > > >> >> >>>> is Avro >> > > >> >> >>>>> > > encoded. >> > > >> >> >>>>> > > > >> What if someone wants to use JSON >> > instead >> > > for a >> > > >> >> good >> > > >> >> >>>> reason? >> > > >> >> >>>>> > What if >> > > >> >> >>>>> > > > user X >> > > >> >> >>>>> > > > >> wants to encrypt messages, but user= Y >> > > does not? >> > > >> >> >>>> Maintaining the >> > > >> >> >>>>> > client >> > > >> >> >>>>> > > > >> library is still much easier than >> > > maintaining the >> > > >> >> >>>> message >> > > >> >> >>>>> > formats. >> > > >> >> >>>>> > > > >> >> > > >> >> >>>>> > > > >> -Todd >> > > >> >> >>>>> > > > >> >> > > >> >> >>>>> > > > >> >> > > >> >> >>>>> > > > >> On Thu, Dec 1, 2016 at 6:21 PM, Gwe= n >> > > Shapira < >> > > >> >> >>>> gwen@confluent.io >> > > >> >> >>>>> > > >> > > >> >> >>>>> > > wrote: >> > > >> >> >>>>> > > > >> >> > > >> >> >>>>> > > > >> > Based on your last sentence, >> consider >> > me >> > > >> >> convinced >> > > >> >> >> :) >> > > >> >> >>>>> > > > >> > >> > > >> >> >>>>> > > > >> > I get why headers are critical fo= r >> > > Mirroring >> > > >> (you >> > > >> >> >>>> need tags to >> > > >> >> >>>>> > > prevent >> > > >> >> >>>>> > > > >> > loops and sometimes to route >> messages >> > > to the >> > > >> >> correct >> > > >> >> >>>>> > destination). >> > > >> >> >>>>> > > > >> > But why do you need headers to >> audit? >> > > We are >> > > >> >> >> auditing >> > > >> >> >>>> by >> > > >> >> >>>>> > producing >> > > >> >> >>>>> > > > >> > counts to a side topic (and I was >> > under >> > > the >> > > >> >> >>>> impression you do >> > > >> >> >>>>> > the >> > > >> >> >>>>> > > > >> > same), so we never need to modify >> the >> > > message. >> > > >> >> >>>>> > > > >> > >> > > >> >> >>>>> > > > >> > Another thing - after we added >> > headers, >> > > >> wouldn't >> > > >> >> you >> > > >> >> >>>> be in the >> > > >> >> >>>>> > > > >> > business of making sure everyone >> uses >> > > them >> > > >> >> properly? >> > > >> >> >>>> Making >> > > >> >> >>>>> > sure >> > > >> >> >>>>> > > > >> > everyone includes the right heade= rs >> > you >> > > need, >> > > >> not >> > > >> >> >>>> using the >> > > >> >> >>>>> > header >> > > >> >> >>>>> > > > >> > names you intend to use, etc. I >> don't >> > > think the >> > > >> >> >>>> "policing" >> > > >> >> >>>>> > business >> > > >> >> >>>>> > > > >> > will ever go away. >> > > >> >> >>>>> > > > >> > >> > > >> >> >>>>> > > > >> > On Thu, Dec 1, 2016 at 5:25 PM, T= odd >> > > Palino < >> > > >> >> >>>>> > tpalino@gmail.com> >> > > >> >> >>>>> > > > wrote: >> > > >> >> >>>>> > > > >> > > Got it. As an ops guy, I'm not >> very >> > > happy >> > > >> with >> > > >> >> the >> > > >> >> >>>>> > workaround. >> > > >> >> >>>>> > > Avro >> > > >> >> >>>>> > > > >> means >> > > >> >> >>>>> > > > >> > > that I have to be concerned wit= h >> the >> > > format >> > > >> of >> > > >> >> the >> > > >> >> >>>> messages >> > > >> >> >>>>> > in >> > > >> >> >>>>> > > > order to >> > > >> >> >>>>> > > > >> > run >> > > >> >> >>>>> > > > >> > > the infrastructure (audit, >> > mirroring, >> > > etc.). >> > > >> >> That >> > > >> >> >>>> means >> > > >> >> >>>>> > that I >> > > >> >> >>>>> > > have >> > > >> >> >>>>> > > > to >> > > >> >> >>>>> > > > >> > > handle the schemas, and I have = to >> > > enforce >> > > >> rules >> > > >> >> >>>> about good >> > > >> >> >>>>> > > formats. >> > > >> >> >>>>> > > > >> This >> > > >> >> >>>>> > > > >> > is >> > > >> >> >>>>> > > > >> > > not something I want to be in t= he >> > > business >> > > >> of, >> > > >> >> >>>> because I >> > > >> >> >>>>> > should be >> > > >> >> >>>>> > > > able >> > > >> >> >>>>> > > > >> > to >> > > >> >> >>>>> > > > >> > > run a service infrastructure >> without >> > > needing >> > > >> >> to be >> > > >> >> >>>> in the >> > > >> >> >>>>> > weeds of >> > > >> >> >>>>> > > > >> > dealing >> > > >> >> >>>>> > > > >> > > with customer data formats. >> > > >> >> >>>>> > > > >> > > >> > > >> >> >>>>> > > > >> > > Trust me, a sizable portion of = my >> > > support >> > > >> time >> > > >> >> is >> > > >> >> >>>> spent >> > > >> >> >>>>> > dealing >> > > >> >> >>>>> > > with >> > > >> >> >>>>> > > > >> > schema >> > > >> >> >>>>> > > > >> > > issues. I really would like to = get >> > > away from >> > > >> >> that. >> > > >> >> >>>> Maybe >> > > >> >> >>>>> > I'd have >> > > >> >> >>>>> > > > more >> > > >> >> >>>>> > > > >> > time >> > > >> >> >>>>> > > > >> > > for other hobbies. Like writing= . >> ;) >> > > >> >> >>>>> > > > >> > > >> > > >> >> >>>>> > > > >> > > -Todd >> > > >> >> >>>>> > > > >> > > >> > > >> >> >>>>> > > > >> > > On Thu, Dec 1, 2016 at 4:04 PM >> Gwen >> > > Shapira < >> > > >> >> >>>>> > gwen@confluent.io> >> > > >> >> >>>>> > > > wrote: >> > > >> >> >>>>> > > > >> > > >> > > >> >> >>>>> > > > >> > >> I'm pretty satisfied with the >> > current >> > > >> >> workarounds >> > > >> >> >>>> (Avro >> > > >> >> >>>>> > container >> > > >> >> >>>>> > > > >> > >> format), so I'm not too excite= d >> > > about the >> > > >> >> extra >> > > >> >> >>>> work >> > > >> >> >>>>> > required to >> > > >> >> >>>>> > > do >> > > >> >> >>>>> > > > >> > >> headers in Kafka. I absolutely >> > don't >> > > mind it >> > > >> >> if >> > > >> >> >>>> you do >> > > >> >> >>>>> > it... >> > > >> >> >>>>> > > > >> > >> I think the Apache convention = for >> > > "good >> > > >> idea, >> > > >> >> but >> > > >> >> >>>> not >> > > >> >> >>>>> > willing to >> > > >> >> >>>>> > > > put >> > > >> >> >>>>> > > > >> > >> any work toward it" is +0.5? >> > anyway, >> > > that's >> > > >> >> what >> > > >> >> >> I >> > > >> >> >>>> was >> > > >> >> >>>>> > trying to >> > > >> >> >>>>> > > > >> > >> convey :) >> > > >> >> >>>>> > > > >> > >> >> > > >> >> >>>>> > > > >> > >> On Thu, Dec 1, 2016 at 3:05 PM= , >> > Todd >> > > Palino >> > > >> < >> > > >> >> >>>>> > tpalino@gmail.com> >> > > >> >> >>>>> > > > >> wrote: >> > > >> >> >>>>> > > > >> > >> > Well I guess my question for >> you, >> > > then, is >> > > >> >> what >> > > >> >> >>>> is >> > > >> >> >>>>> > holding you >> > > >> >> >>>>> > > > back >> > > >> >> >>>>> > > > >> > from >> > > >> >> >>>>> > > > >> > >> > full support for headers? >> What=E2=80=99s >> > > the bit >> > > >> >> that >> > > >> >> >>>> you=E2=80=99re >> > > >> >> >>>>> > missing >> > > >> >> >>>>> > > that >> > > >> >> >>>>> > > > >> has >> > > >> >> >>>>> > > > >> > you >> > > >> >> >>>>> > > > >> > >> > under a full +1? >> > > >> >> >>>>> > > > >> > >> > >> > > >> >> >>>>> > > > >> > >> > -Todd >> > > >> >> >>>>> > > > >> > >> > >> > > >> >> >>>>> > > > >> > >> > >> > > >> >> >>>>> > > > >> > >> > On Thu, Dec 1, 2016 at 1:59 = PM, >> > > Gwen >> > > >> >> Shapira < >> > > >> >> >>>>> > > gwen@confluent.io> >> > > >> >> >>>>> > > > >> > wrote: >> > > >> >> >>>>> > > > >> > >> > >> > > >> >> >>>>> > > > >> > >> >> I know why people who suppo= rt >> > > headers >> > > >> >> support >> > > >> >> >>>> them, and >> > > >> >> >>>>> > I've >> > > >> >> >>>>> > > > seen >> > > >> >> >>>>> > > > >> > what >> > > >> >> >>>>> > > > >> > >> >> the discussion is like. >> > > >> >> >>>>> > > > >> > >> >> >> > > >> >> >>>>> > > > >> > >> >> This is why I'm asking peop= le >> > who >> > > are >> > > >> >> against >> > > >> >> >>>> headers >> > > >> >> >>>>> > > > (especially >> > > >> >> >>>>> > > > >> > >> >> committers) what will make >> them >> > > change >> > > >> >> their >> > > >> >> >>>> mind - so >> > > >> >> >>>>> > we can >> > > >> >> >>>>> > > > get >> > > >> >> >>>>> > > > >> > this >> > > >> >> >>>>> > > > >> > >> >> part over one way or anothe= r. >> > > >> >> >>>>> > > > >> > >> >> >> > > >> >> >>>>> > > > >> > >> >> If I sound frustrated it is >> not >> > > at Radai, >> > > >> >> Jun >> > > >> >> >>>> or you >> > > >> >> >>>>> > (Todd)... >> > > >> >> >>>>> > > > I am >> > > >> >> >>>>> > > > >> > >> >> just looking for something >> > > concrete we >> > > >> can >> > > >> >> do >> > > >> >> >>>> to move >> > > >> >> >>>>> > the >> > > >> >> >>>>> > > > >> discussion >> > > >> >> >>>>> > > > >> > >> >> along to the yummy design >> > details >> > > (which >> > > >> is >> > > >> >> >> the >> > > >> >> >>>>> > argument I >> > > >> >> >>>>> > > > really >> > > >> >> >>>>> > > > >> am >> > > >> >> >>>>> > > > >> > >> >> looking forward to). >> > > >> >> >>>>> > > > >> > >> >> >> > > >> >> >>>>> > > > >> > >> >> On Thu, Dec 1, 2016 at 1:53 >> PM, >> > > Todd >> > > >> >> Palino < >> > > >> >> >>>>> > > tpalino@gmail.com> >> > > >> >> >>>>> > > > >> > wrote: >> > > >> >> >>>>> > > > >> > >> >> > So, Gwen, to your questio= n >> > > (even though >> > > >> >> I=E2=80=99m >> > > >> >> >>>> not a >> > > >> >> >>>>> > > > committer)... >> > > >> >> >>>>> > > > >> > >> >> > >> > > >> >> >>>>> > > > >> > >> >> > I have always been a stro= ng >> > > supporter >> > > >> of >> > > >> >> >>>> introducing >> > > >> >> >>>>> > the >> > > >> >> >>>>> > > > concept >> > > >> >> >>>>> > > > >> > of an >> > > >> >> >>>>> > > > >> > >> >> > envelope to messages, whi= ch >> > > headers >> > > >> >> >>>> accomplishes. The >> > > >> >> >>>>> > > message >> > > >> >> >>>>> > > > key >> > > >> >> >>>>> > > > >> > is >> > > >> >> >>>>> > > > >> > >> >> > already an example of a >> piece >> > of >> > > >> envelope >> > > >> >> >>>>> > information. By >> > > >> >> >>>>> > > > >> > providing a >> > > >> >> >>>>> > > > >> > >> >> means >> > > >> >> >>>>> > > > >> > >> >> > to do this within Kafka >> > itself, >> > > and not >> > > >> >> >>>> relying on >> > > >> >> >>>>> > use-case >> > > >> >> >>>>> > > > >> > specific >> > > >> >> >>>>> > > > >> > >> >> > implementations, you make= it >> > > much >> > > >> easier >> > > >> >> for >> > > >> >> >>>>> > components to >> > > >> >> >>>>> > > > >> > >> interoperate. >> > > >> >> >>>>> > > > >> > >> >> It >> > > >> >> >>>>> > > > >> > >> >> > simplifies development of >> all >> > > these >> > > >> >> things >> > > >> >> >>>> (message >> > > >> >> >>>>> > routing, >> > > >> >> >>>>> > > > >> > auditing, >> > > >> >> >>>>> > > > >> > >> >> > encryption, etc.) because >> each >> > > one does >> > > >> >> not >> > > >> >> >>>> have to >> > > >> >> >>>>> > reinvent >> > > >> >> >>>>> > > > the >> > > >> >> >>>>> > > > >> > >> wheel. >> > > >> >> >>>>> > > > >> > >> >> > >> > > >> >> >>>>> > > > >> > >> >> > It also makes it much eas= ier >> > > from a >> > > >> >> client >> > > >> >> >>>> point of >> > > >> >> >>>>> > view if >> > > >> >> >>>>> > > > the >> > > >> >> >>>>> > > > >> > >> headers >> > > >> >> >>>>> > > > >> > >> >> are >> > > >> >> >>>>> > > > >> > >> >> > defined as part of the >> > protocol >> > > and/or >> > > >> >> >>>> message format >> > > >> >> >>>>> > in >> > > >> >> >>>>> > > > general >> > > >> >> >>>>> > > > >> > >> because >> > > >> >> >>>>> > > > >> > >> >> > you can easily produce an= d >> > > consume >> > > >> >> messages >> > > >> >> >>>> without >> > > >> >> >>>>> > having >> > > >> >> >>>>> > > to >> > > >> >> >>>>> > > > >> take >> > > >> >> >>>>> > > > >> > >> into >> > > >> >> >>>>> > > > >> > >> >> > account specific cases. F= or >> > > example, I >> > > >> >> want >> > > >> >> >>>> to route >> > > >> >> >>>>> > > messages, >> > > >> >> >>>>> > > > >> but >> > > >> >> >>>>> > > > >> > >> >> client A >> > > >> >> >>>>> > > > >> > >> >> > doesn=E2=80=99t support t= he way >> audit >> > > >> implemented >> > > >> >> >>>> headers, and >> > > >> >> >>>>> > > client >> > > >> >> >>>>> > > > B >> > > >> >> >>>>> > > > >> > >> doesn=E2=80=99t >> > > >> >> >>>>> > > > >> > >> >> > support the way encryptio= n >> or >> > > routing >> > > >> >> >>>> implemented >> > > >> >> >>>>> > headers, >> > > >> >> >>>>> > > so >> > > >> >> >>>>> > > > now >> > > >> >> >>>>> > > > >> > my >> > > >> >> >>>>> > > > >> > >> >> > application has to create >> some >> > > really >> > > >> >> >> fragile >> > > >> >> >>>> (my >> > > >> >> >>>>> > > autocorrect >> > > >> >> >>>>> > > > >> just >> > > >> >> >>>>> > > > >> > >> tried >> > > >> >> >>>>> > > > >> > >> >> to >> > > >> >> >>>>> > > > >> > >> >> > make that =E2=80=9Ctragic= =E2=80=9D, which is >> > > probably >> > > >> >> >>>> appropriate >> > > >> >> >>>>> > too) code >> > > >> >> >>>>> > > to >> > > >> >> >>>>> > > > >> > strip >> > > >> >> >>>>> > > > >> > >> >> > everything off, rather th= an >> > just >> > > >> >> consuming >> > > >> >> >> the >> > > >> >> >>>>> > messages, >> > > >> >> >>>>> > > > picking >> > > >> >> >>>>> > > > >> > out >> > > >> >> >>>>> > > > >> > >> the >> > > >> >> >>>>> > > > >> > >> >> 1 >> > > >> >> >>>>> > > > >> > >> >> > or 2 headers it=E2=80=99s= interested >> > > in, and >> > > >> >> >>>> performing its >> > > >> >> >>>>> > > function. >> > > >> >> >>>>> > > > >> > >> >> > >> > > >> >> >>>>> > > > >> > >> >> > Honestly, this discussion >> has >> > > been >> > > >> going >> > > >> >> on >> > > >> >> >>>> for a >> > > >> >> >>>>> > long time, >> > > >> >> >>>>> > > > and >> > > >> >> >>>>> > > > >> > it=E2=80=99s >> > > >> >> >>>>> > > > >> > >> >> > always =E2=80=9COh, you c= ame up >> with 2 >> > > use >> > > >> cases, >> > > >> >> >> and >> > > >> >> >>>> yeah, >> > > >> >> >>>>> > those >> > > >> >> >>>>> > > use >> > > >> >> >>>>> > > > >> cases >> > > >> >> >>>>> > > > >> > >> are >> > > >> >> >>>>> > > > >> > >> >> > real things that someone >> would >> > > want to >> > > >> >> do. >> > > >> >> >>>> Here=E2=80=99s an >> > > >> >> >>>>> > > alternate >> > > >> >> >>>>> > > > >> way >> > > >> >> >>>>> > > > >> > to >> > > >> >> >>>>> > > > >> > >> >> > implement them so let=E2= =80=99s not >> do >> > > >> headers.=E2=80=9D >> > > >> >> If >> > > >> >> >>>> we have a >> > > >> >> >>>>> > few >> > > >> >> >>>>> > > use >> > > >> >> >>>>> > > > >> cases >> > > >> >> >>>>> > > > >> > >> that >> > > >> >> >>>>> > > > >> > >> >> we >> > > >> >> >>>>> > > > >> > >> >> > actually came up with, yo= u >> can >> > > be sure >> > > >> >> that >> > > >> >> >>>> over the >> > > >> >> >>>>> > next >> > > >> >> >>>>> > > year >> > > >> >> >>>>> > > > >> > >> there=E2=80=99s a >> > > >> >> >>>>> > > > >> > >> >> > dozen others that we didn= =E2=80=99t >> > > think of >> > > >> that >> > > >> >> >>>> someone >> > > >> >> >>>>> > would like >> > > >> >> >>>>> > > > to >> > > >> >> >>>>> > > > >> > do. I >> > > >> >> >>>>> > > > >> > >> >> > really think it=E2=80=99s= time to >> stop >> > > >> rehashing >> > > >> >> >> this >> > > >> >> >>>>> > discussion and >> > > >> >> >>>>> > > > >> > instead >> > > >> >> >>>>> > > > >> > >> >> focus >> > > >> >> >>>>> > > > >> > >> >> > on a workable standard th= at >> we >> > > can >> > > >> adopt. >> > > >> >> >>>>> > > > >> > >> >> > >> > > >> >> >>>>> > > > >> > >> >> > -Todd >> > > >> >> >>>>> > > > >> > >> >> > >> > > >> >> >>>>> > > > >> > >> >> > >> > > >> >> >>>>> > > > >> > >> >> > On Thu, Dec 1, 2016 at 1:= 39 >> > PM, >> > > Todd >> > > >> >> Palino >> > > >> >> >> < >> > > >> >> >>>>> > > > tpalino@gmail.com> >> > > >> >> >>>>> > > > >> > >> wrote: >> > > >> >> >>>>> > > > >> > >> >> > >> > > >> >> >>>>> > > > >> > >> >> >> C. per message encryptio= n >> > > >> >> >>>>> > > > >> > >> >> >>> One drawback of this >> > approach >> > > is that >> > > >> >> this >> > > >> >> >>>>> > significantly >> > > >> >> >>>>> > > > reduce >> > > >> >> >>>>> > > > >> > the >> > > >> >> >>>>> > > > >> > >> >> >>> effectiveness of >> > compression, >> > > which >> > > >> >> >> happens >> > > >> >> >>>> on a >> > > >> >> >>>>> > set of >> > > >> >> >>>>> > > > >> > serialized >> > > >> >> >>>>> > > > >> > >> >> >>> messages. An alternativ= e >> is >> > > to enable >> > > >> >> SSL >> > > >> >> >>>> for wire >> > > >> >> >>>>> > > > encryption >> > > >> >> >>>>> > > > >> and >> > > >> >> >>>>> > > > >> > >> rely >> > > >> >> >>>>> > > > >> > >> >> on >> > > >> >> >>>>> > > > >> > >> >> >>> the storage system (e.g= . >> > > LUKS) for at >> > > >> >> rest >> > > >> >> >>>>> > encryption. >> > > >> >> >>>>> > > > >> > >> >> >> >> > > >> >> >>>>> > > > >> > >> >> >> >> > > >> >> >>>>> > > > >> > >> >> >> Jun, this is not >> sufficient. >> > > While >> > > >> this >> > > >> >> >> does >> > > >> >> >>>> cover >> > > >> >> >>>>> > the case >> > > >> >> >>>>> > > > of >> > > >> >> >>>>> > > > >> > >> removing >> > > >> >> >>>>> > > > >> > >> >> a >> > > >> >> >>>>> > > > >> > >> >> >> drive from the system, i= t >> > will >> > > not >> > > >> >> satisfy >> > > >> >> >>>> most >> > > >> >> >>>>> > compliance >> > > >> >> >>>>> > > > >> > >> requirements >> > > >> >> >>>>> > > > >> > >> >> for >> > > >> >> >>>>> > > > >> > >> >> >> encryption of data as >> whoever >> > > has >> > > >> >> access to >> > > >> >> >>>> the >> > > >> >> >>>>> > broker >> > > >> >> >>>>> > > itself >> > > >> >> >>>>> > > > >> > still >> > > >> >> >>>>> > > > >> > >> has >> > > >> >> >>>>> > > > >> > >> >> >> access to the unencrypte= d >> > > data. For >> > > >> >> >>>> end-to-end >> > > >> >> >>>>> > encryption >> > > >> >> >>>>> > > you >> > > >> >> >>>>> > > > >> > need to >> > > >> >> >>>>> > > > >> > >> >> >> encrypt at the producer, >> > > before it >> > > >> >> enters >> > > >> >> >> the >> > > >> >> >>>>> > system, and >> > > >> >> >>>>> > > > >> decrypt >> > > >> >> >>>>> > > > >> > at >> > > >> >> >>>>> > > > >> > >> the >> > > >> >> >>>>> > > > >> > >> >> >> consumer, after it exits >> the >> > > system. >> > > >> >> >>>>> > > > >> > >> >> >> >> > > >> >> >>>>> > > > >> > >> >> >> -Todd >> > > >> >> >>>>> > > > >> > >> >> >> >> > > >> >> >>>>> > > > >> > >> >> >> >> > > >> >> >>>>> > > > >> > >> >> >> On Thu, Dec 1, 2016 at 1= :03 >> > > PM, radai >> > > >> < >> > > >> >> >>>>> > > > >> radai.rosenblatt@gmail.com >> > > >> >> >>>>> > > > >> > > >> > > >> >> >>>>> > > > >> > >> >> wrote: >> > > >> >> >>>>> > > > >> > >> >> >> >> > > >> >> >>>>> > > > >> > >> >> >>> another big plus of >> headers >> > > in the >> > > >> >> >> protocol >> > > >> >> >>>> is that >> > > >> >> >>>>> > it >> > > >> >> >>>>> > > would >> > > >> >> >>>>> > > > >> > enable >> > > >> >> >>>>> > > > >> > >> >> rapid >> > > >> >> >>>>> > > > >> > >> >> >>> iteration on ideas outs= ide >> > of >> > > core >> > > >> >> kafka >> > > >> >> >>>> and would >> > > >> >> >>>>> > reduce >> > > >> >> >>>>> > > > the >> > > >> >> >>>>> > > > >> > >> number of >> > > >> >> >>>>> > > > >> > >> >> >>> future wire format chan= ges >> > > required. >> > > >> >> >>>>> > > > >> > >> >> >>> >> > > >> >> >>>>> > > > >> > >> >> >>> a lot of what is >> currently a >> > > KIP >> > > >> >> >> represents >> > > >> >> >>>> use >> > > >> >> >>>>> > cases that >> > > >> >> >>>>> > > > are >> > > >> >> >>>>> > > > >> > not >> > > >> >> >>>>> > > > >> > >> 100% >> > > >> >> >>>>> > > > >> > >> >> >>> relevant to all users, = and >> > > some of >> > > >> them >> > > >> >> >>>> require >> > > >> >> >>>>> > rather >> > > >> >> >>>>> > > > invasive >> > > >> >> >>>>> > > > >> > wire >> > > >> >> >>>>> > > > >> > >> >> >>> protocol changes. a thi= ng >> a >> > > good >> > > >> recent >> > > >> >> >>>> example of >> > > >> >> >>>>> > this is >> > > >> >> >>>>> > > > >> > kip-98. >> > > >> >> >>>>> > > > >> > >> >> >>> tx-utilizing traffic is >> > > expected to >> > > >> be >> > > >> >> a >> > > >> >> >>>> very small >> > > >> >> >>>>> > > > fraction of >> > > >> >> >>>>> > > > >> > >> total >> > > >> >> >>>>> > > > >> > >> >> >>> traffic and yet the >> changes >> > > are >> > > >> >> invasive. >> > > >> >> >>>>> > > > >> > >> >> >>> >> > > >> >> >>>>> > > > >> > >> >> >>> every such wire format >> > change >> > > >> >> translates >> > > >> >> >>>> into >> > > >> >> >>>>> > painful and >> > > >> >> >>>>> > > > slow >> > > >> >> >>>>> > > > >> > >> >> adoption of >> > > >> >> >>>>> > > > >> > >> >> >>> new versions. >> > > >> >> >>>>> > > > >> > >> >> >>> >> > > >> >> >>>>> > > > >> > >> >> >>> i think a lot of >> > functionality >> > > >> >> currently >> > > >> >> >> in >> > > >> >> >>>> KIPs >> > > >> >> >>>>> > could be >> > > >> >> >>>>> > > > "spun >> > > >> >> >>>>> > > > >> > out" >> > > >> >> >>>>> > > > >> > >> >> and >> > > >> >> >>>>> > > > >> > >> >> >>> implemented as opt-in >> > plugins >> > > >> >> transmitting >> > > >> >> >>>> data over >> > > >> >> >>>>> > > > headers. >> > > >> >> >>>>> > > > >> > this >> > > >> >> >>>>> > > > >> > >> >> would >> > > >> >> >>>>> > > > >> > >> >> >>> keep the core wire form= at >> > > stable(r), >> > > >> >> core >> > > >> >> >>>> codebase >> > > >> >> >>>>> > > smaller, >> > > >> >> >>>>> > > > and >> > > >> >> >>>>> > > > >> > >> avoid >> > > >> >> >>>>> > > > >> > >> >> the >> > > >> >> >>>>> > > > >> > >> >> >>> "burden of proof" thats >> > > sometimes >> > > >> >> required >> > > >> >> >>>> to prove >> > > >> >> >>>>> > a >> > > >> >> >>>>> > > > certain >> > > >> >> >>>>> > > > >> > >> feature >> > > >> >> >>>>> > > > >> > >> >> is >> > > >> >> >>>>> > > > >> > >> >> >>> useful enough for a >> > > wide-enough >> > > >> >> audience >> > > >> >> >> to >> > > >> >> >>>> warrant >> > > >> >> >>>>> > a wire >> > > >> >> >>>>> > > > >> format >> > > >> >> >>>>> > > > >> > >> >> change >> > > >> >> >>>>> > > > >> > >> >> >>> and code complexity >> > additions. >> > > >> >> >>>>> > > > >> > >> >> >>> >> > > >> >> >>>>> > > > >> > >> >> >>> (to be clear - kip-98 g= oes >> > > beyond >> > > >> >> "mere" >> > > >> >> >>>> wire format >> > > >> >> >>>>> > > changes >> > > >> >> >>>>> > > > >> and >> > > >> >> >>>>> > > > >> > im >> > > >> >> >>>>> > > > >> > >> not >> > > >> >> >>>>> > > > >> > >> >> >>> saying it could have be= en >> > > completely >> > > >> >> done >> > > >> >> >>>> with >> > > >> >> >>>>> > headers, >> > > >> >> >>>>> > > but >> > > >> >> >>>>> > > > >> > >> >> exactly-once >> > > >> >> >>>>> > > > >> > >> >> >>> delivery certainly coul= d) >> > > >> >> >>>>> > > > >> > >> >> >>> >> > > >> >> >>>>> > > > >> > >> >> >>> On Thu, Dec 1, 2016 at >> 11:20 >> > > AM, Gwen >> > > >> >> >>>> Shapira < >> > > >> >> >>>>> > > > >> gwen@confluent.io >> > > >> >> >>>>> > > > >> > > >> > > >> >> >>>>> > > > >> > >> >> wrote: >> > > >> >> >>>>> > > > >> > >> >> >>> >> > > >> >> >>>>> > > > >> > >> >> >>> > On Thu, Dec 1, 2016 a= t >> > > 10:24 AM, >> > > >> >> radai < >> > > >> >> >>>>> > > > >> > >> radai.rosenblatt@gmail.com> >> > > >> >> >>>>> > > > >> > >> >> >>> wrote: >> > > >> >> >>>>> > > > >> > >> >> >>> > > "For use cases with= in >> an >> > > >> >> organization, >> > > >> >> >>>> one could >> > > >> >> >>>>> > > always >> > > >> >> >>>>> > > > use >> > > >> >> >>>>> > > > >> > >> other >> > > >> >> >>>>> > > > >> > >> >> >>> > > approaches such as >> > > company-wise >> > > >> >> >>>> containers" >> > > >> >> >>>>> > > > >> > >> >> >>> > > this is what linked= in >> > has >> > > >> >> >> traditionally >> > > >> >> >>>> done >> > > >> >> >>>>> > but there >> > > >> >> >>>>> > > > are >> > > >> >> >>>>> > > > >> > now >> > > >> >> >>>>> > > > >> > >> >> cases >> > > >> >> >>>>> > > > >> > >> >> >>> > (read >> > > >> >> >>>>> > > > >> > >> >> >>> > > - topics) where thi= s >> is >> > > not >> > > >> >> >> acceptable. >> > > >> >> >>>> this >> > > >> >> >>>>> > makes >> > > >> >> >>>>> > > > headers >> > > >> >> >>>>> > > > >> > >> useful >> > > >> >> >>>>> > > > >> > >> >> even >> > > >> >> >>>>> > > > >> > >> >> >>> > > within single orgs = for >> > > cases >> > > >> where >> > > >> >> >>>>> > > > one-container-fits-all >> > > >> >> >>>>> > > > >> > cannot >> > > >> >> >>>>> > > > >> > >> >> >>> apply. >> > > >> >> >>>>> > > > >> > >> >> >>> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > > as for the particul= ar >> > use >> > > cases >> > > >> >> >> listed, >> > > >> >> >>>> i dont >> > > >> >> >>>>> > want >> > > >> >> >>>>> > > > this to >> > > >> >> >>>>> > > > >> > >> devolve >> > > >> >> >>>>> > > > >> > >> >> >>> to a >> > > >> >> >>>>> > > > >> > >> >> >>> > > discussion of >> particular >> > > use >> > > >> cases >> > > >> >> - i >> > > >> >> >>>> think its >> > > >> >> >>>>> > > enough >> > > >> >> >>>>> > > > >> that >> > > >> >> >>>>> > > > >> > >> some >> > > >> >> >>>>> > > > >> > >> >> of >> > > >> >> >>>>> > > > >> > >> >> >>> them >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> >>>>> > > > >> > >> >> >>> > I think a main point = of >> > > contention >> > > >> is >> > > >> >> >>>> that: We >> > > >> >> >>>>> > > identified >> > > >> >> >>>>> > > > few >> > > >> >> >>>>> > > > >> > >> >> >>> > use-cases where heade= rs >> > are >> > > useful, >> > > >> >> do >> > > >> >> >> we >> > > >> >> >>>> want >> > > >> >> >>>>> > Kafka to >> > > >> >> >>>>> > > > be a >> > > >> >> >>>>> > > > >> > >> system >> > > >> >> >>>>> > > > >> > >> >> >>> > that supports those >> > > use-cases? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> >>>>> > > > >> > >> >> >>> > For example, Jun said= : >> > > >> >> >>>>> > > > >> > >> >> >>> > "Not sure how widely >> > useful >> > > >> >> record-level >> > > >> >> >>>> lineage >> > > >> >> >>>>> > is >> > > >> >> >>>>> > > though >> > > >> >> >>>>> > > > >> > since >> > > >> >> >>>>> > > > >> > >> the >> > > >> >> >>>>> > > > >> > >> >> >>> > overhead could >> > > >> >> >>>>> > > > >> > >> >> >>> > be significant." >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> >>>>> > > > >> > >> >> >>> > We know NiFi supports >> > > record level >> > > >> >> >>>> lineage. I >> > > >> >> >>>>> > don't >> > > >> >> >>>>> > > think >> > > >> >> >>>>> > > > it >> > > >> >> >>>>> > > > >> > was >> > > >> >> >>>>> > > > >> > >> >> >>> > developed for lols, I >> > think >> > > it is >> > > >> >> safe >> > > >> >> >> to >> > > >> >> >>>> assume >> > > >> >> >>>>> > that >> > > >> >> >>>>> > > the >> > > >> >> >>>>> > > > NSA >> > > >> >> >>>>> > > > >> > >> needed >> > > >> >> >>>>> > > > >> > >> >> >>> > that functionality. W= e >> > also >> > > know >> > > >> that >> > > >> >> >>>> certain >> > > >> >> >>>>> > financial >> > > >> >> >>>>> > > > >> > institutes >> > > >> >> >>>>> > > > >> > >> >> >>> > need to track tamperi= ng >> > with >> > > >> records >> > > >> >> at >> > > >> >> >> a >> > > >> >> >>>> record >> > > >> >> >>>>> > level >> > > >> >> >>>>> > > and >> > > >> >> >>>>> > > > >> > there >> > > >> >> >>>>> > > > >> > >> are >> > > >> >> >>>>> > > > >> > >> >> >>> > federal regulations t= hat >> > > absolutely >> > > >> >> >>>> require >> > > >> >> >>>>> > this. They >> > > >> >> >>>>> > > > also >> > > >> >> >>>>> > > > >> > need >> > > >> >> >>>>> > > > >> > >> to >> > > >> >> >>>>> > > > >> > >> >> >>> > prove that routing ap= ps >> > that >> > > >> >> "touches" >> > > >> >> >> the >> > > >> >> >>>>> > messages and >> > > >> >> >>>>> > > > >> either >> > > >> >> >>>>> > > > >> > >> reads >> > > >> >> >>>>> > > > >> > >> >> >>> > or updates headers >> > couldn't >> > > have >> > > >> >> >> possibly >> > > >> >> >>>>> > modified the >> > > >> >> >>>>> > > > >> payload >> > > >> >> >>>>> > > > >> > >> >> itself. >> > > >> >> >>>>> > > > >> > >> >> >>> > They use record level >> > > encryption to >> > > >> >> do >> > > >> >> >>>> that - >> > > >> >> >>>>> > apps can >> > > >> >> >>>>> > > > read >> > > >> >> >>>>> > > > >> and >> > > >> >> >>>>> > > > >> > >> >> >>> > (sometimes) modify >> headers >> > > but >> > > >> can't >> > > >> >> >>>> touch the >> > > >> >> >>>>> > payload. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> >>>>> > > > >> > >> >> >>> > We can totally say >> "those >> > > are >> > > >> corner >> > > >> >> >>>> cases and >> > > >> >> >>>>> > not worth >> > > >> >> >>>>> > > > >> adding >> > > >> >> >>>>> > > > >> > >> >> >>> > headers to Kafka for"= , >> > they >> > > should >> > > >> >> use a >> > > >> >> >>>> different >> > > >> >> >>>>> > > pubsub >> > > >> >> >>>>> > > > >> > message >> > > >> >> >>>>> > > > >> > >> for >> > > >> >> >>>>> > > > >> > >> >> >>> > that (Nifi or one of = the >> > > other 1000 >> > > >> >> that >> > > >> >> >>>> cater >> > > >> >> >>>>> > > > specifically >> > > >> >> >>>>> > > > >> to >> > > >> >> >>>>> > > > >> > the >> > > >> >> >>>>> > > > >> > >> >> >>> > financial industry). >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> >>>>> > > > >> > >> >> >>> > But this gets us into= a >> > > catch 22: >> > > >> >> >>>>> > > > >> > >> >> >>> > If we discuss a speci= fic >> > > use-case, >> > > >> >> >>>> someone can >> > > >> >> >>>>> > always >> > > >> >> >>>>> > > say >> > > >> >> >>>>> > > > it >> > > >> >> >>>>> > > > >> > isn't >> > > >> >> >>>>> > > > >> > >> >> >>> > interesting enough fo= r >> > > Kafka. If we >> > > >> >> >>>> discuss more >> > > >> >> >>>>> > general >> > > >> >> >>>>> > > > >> > trends, >> > > >> >> >>>>> > > > >> > >> >> >>> > others can say "well,= we >> > > are not >> > > >> sure >> > > >> >> >> any >> > > >> >> >>>> of them >> > > >> >> >>>>> > really >> > > >> >> >>>>> > > > >> needs >> > > >> >> >>>>> > > > >> > >> >> headers >> > > >> >> >>>>> > > > >> > >> >> >>> > specifically. This is >> just >> > > hand >> > > >> >> waving >> > > >> >> >>>> and not >> > > >> >> >>>>> > > > interesting.". >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> >>>>> > > > >> > >> >> >>> > I think discussing >> > > use-cases in >> > > >> >> >> specifics >> > > >> >> >>>> is super >> > > >> >> >>>>> > > > important >> > > >> >> >>>>> > > > >> to >> > > >> >> >>>>> > > > >> > >> >> decide >> > > >> >> >>>>> > > > >> > >> >> >>> > implementation detail= s >> for >> > > headers >> > > >> >> (my >> > > >> >> >>>> use-cases >> > > >> >> >>>>> > lean >> > > >> >> >>>>> > > > toward >> > > >> >> >>>>> > > > >> > >> >> numerical >> > > >> >> >>>>> > > > >> > >> >> >>> > keys with namespaces = and >> > > object >> > > >> >> values, >> > > >> >> >>>> others >> > > >> >> >>>>> > differ), >> > > >> >> >>>>> > > > but I >> > > >> >> >>>>> > > > >> > >> think >> > > >> >> >>>>> > > > >> > >> >> we >> > > >> >> >>>>> > > > >> > >> >> >>> > need to answer the >> general >> > > "Are we >> > > >> >> going >> > > >> >> >>>> to have >> > > >> >> >>>>> > > headers" >> > > >> >> >>>>> > > > >> > question >> > > >> >> >>>>> > > > >> > >> >> >>> > first. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> >>>>> > > > >> > >> >> >>> > I'd love to hear from >> the >> > > other >> > > >> >> >>>> committers in the >> > > >> >> >>>>> > > > discussion: >> > > >> >> >>>>> > > > >> > >> >> >>> > What would it take to >> > > convince you >> > > >> >> that >> > > >> >> >>>> headers >> > > >> >> >>>>> > in Kafka >> > > >> >> >>>>> > > > are >> > > >> >> >>>>> > > > >> a >> > > >> >> >>>>> > > > >> > >> good >> > > >> >> >>>>> > > > >> > >> >> >>> > idea in general, so w= e >> can >> > > move >> > > >> ahead >> > > >> >> >> and >> > > >> >> >>>> try to >> > > >> >> >>>>> > agree >> > > >> >> >>>>> > > on >> > > >> >> >>>>> > > > the >> > > >> >> >>>>> > > > >> > >> >> details? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> >>>>> > > > >> > >> >> >>> > I feel like we keep >> moving >> > > the goal >> > > >> >> >> posts >> > > >> >> >>>> and >> > > >> >> >>>>> > this is >> > > >> >> >>>>> > > > truly >> > > >> >> >>>>> > > > >> > >> >> exhausting. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> >>>>> > > > >> > >> >> >>> > For the record, I mil= dly >> > > support >> > > >> >> adding >> > > >> >> >>>> headers >> > > >> >> >>>>> > to Kafka >> > > >> >> >>>>> > > > >> > (+0.5?). >> > > >> >> >>>>> > > > >> > >> >> >>> > The community can >> continue >> > > to find >> > > >> >> >>>> workarounds to >> > > >> >> >>>>> > the >> > > >> >> >>>>> > > > issue >> > > >> >> >>>>> > > > >> and >> > > >> >> >>>>> > > > >> > >> there >> > > >> >> >>>>> > > > >> > >> >> >>> > are some benefits to >> > > keeping the >> > > >> >> message >> > > >> >> >>>> format >> > > >> >> >>>>> > and >> > > >> >> >>>>> > > > clients >> > > >> >> >>>>> > > > >> > >> simpler. >> > > >> >> >>>>> > > > >> > >> >> >>> > But I see the usefuln= ess >> > of >> > > headers >> > > >> >> to >> > > >> >> >>>> many >> > > >> >> >>>>> > use-cases >> > > >> >> >>>>> > > and >> > > >> >> >>>>> > > > if >> > > >> >> >>>>> > > > >> we >> > > >> >> >>>>> > > > >> > >> can >> > > >> >> >>>>> > > > >> > >> >> >>> > find a good and >> generally >> > > useful >> > > >> way >> > > >> >> to >> > > >> >> >>>> add it to >> > > >> >> >>>>> > Kafka, >> > > >> >> >>>>> > > > it >> > > >> >> >>>>> > > > >> > will >> > > >> >> >>>>> > > > >> > >> make >> > > >> >> >>>>> > > > >> > >> >> >>> > Kafka easier to use f= or >> > > many - >> > > >> worthy >> > > >> >> >>>> goal in my >> > > >> >> >>>>> > eyes. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> >>>>> > > > >> > >> >> >>> > > are >> > interesting/feasible, >> > > but: >> > > >> >> >>>>> > > > >> > >> >> >>> > > A+B. i think there = are >> > > use cases >> > > >> >> for >> > > >> >> >>>> polyglot >> > > >> >> >>>>> > topics. >> > > >> >> >>>>> > > > >> > >> especially if >> > > >> >> >>>>> > > > >> > >> >> >>> kafka >> > > >> >> >>>>> > > > >> > >> >> >>> > > is being used to >> "trunk" >> > > >> something >> > > >> >> >> else. >> > > >> >> >>>>> > > > >> > >> >> >>> > > D. multiple topics >> would >> > > make it >> > > >> >> >> harder >> > > >> >> >>>> to write >> > > >> >> >>>>> > > > portable >> > > >> >> >>>>> > > > >> > >> consumer >> > > >> >> >>>>> > > > >> > >> >> >>> code. >> > > >> >> >>>>> > > > >> > >> >> >>> > > partition remapping >> > would >> > > mess >> > > >> with >> > > >> >> >>>> locality of >> > > >> >> >>>>> > > > consumption >> > > >> >> >>>>> > > > >> > >> >> >>> guarantees. >> > > >> >> >>>>> > > > >> > >> >> >>> > > E+F. a use case I s= ee >> > for >> > > >> >> >>>> lineage/metadata is >> > > >> >> >>>>> > > > >> > >> billing/chargeback. >> > > >> >> >>>>> > > > >> > >> >> for >> > > >> >> >>>>> > > > >> > >> >> >>> > that >> > > >> >> >>>>> > > > >> > >> >> >>> > > use case it is not >> > enough >> > > to >> > > >> simply >> > > >> >> >>>> record the >> > > >> >> >>>>> > point >> > > >> >> >>>>> > > of >> > > >> >> >>>>> > > > >> > origin, >> > > >> >> >>>>> > > > >> > >> but >> > > >> >> >>>>> > > > >> > >> >> >>> every >> > > >> >> >>>>> > > > >> > >> >> >>> > > replication stop >> (think >> > > mirror >> > > >> >> maker) >> > > >> >> >>>> must also >> > > >> >> >>>>> > add a >> > > >> >> >>>>> > > > >> record >> > > >> >> >>>>> > > > >> > to >> > > >> >> >>>>> > > > >> > >> >> form a >> > > >> >> >>>>> > > > >> > >> >> >>> > > "transit log". >> > > >> >> >>>>> > > > >> > >> >> >>> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > > as for stream >> processing >> > > on top >> > > >> of >> > > >> >> >>>> kafka - i >> > > >> >> >>>>> > know >> > > >> >> >>>>> > > samza >> > > >> >> >>>>> > > > >> has a >> > > >> >> >>>>> > > > >> > >> >> metadata >> > > >> >> >>>>> > > > >> > >> >> >>> > map >> > > >> >> >>>>> > > > >> > >> >> >>> > > which they carry >> around >> > in >> > > >> >> addition to >> > > >> >> >>>> user >> > > >> >> >>>>> > values. >> > > >> >> >>>>> > > > headers >> > > >> >> >>>>> > > > >> > are >> > > >> >> >>>>> > > > >> > >> the >> > > >> >> >>>>> > > > >> > >> >> >>> > perfect >> > > >> >> >>>>> > > > >> > >> >> >>> > > fit for these thing= s. >> > > >> >> >>>>> > > > >> > >> >> >>> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > > On Wed, Nov 30, 201= 6 >> at >> > > 6:50 PM, >> > > >> >> Jun >> > > >> >> >>>> Rao < >> > > >> >> >>>>> > > > jun@confluent.io >> > > >> >> >>>>> > > > >> > >> > > >> >> >>>>> > > > >> > >> wrote: >> > > >> >> >>>>> > > > >> > >> >> >>> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> Hi, Michael, >> > > >> >> >>>>> > > > >> > >> >> >>> > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> In order to answer >> the >> > > first two >> > > >> >> >>>> questions, it >> > > >> >> >>>>> > would >> > > >> >> >>>>> > > be >> > > >> >> >>>>> > > > >> > helpful >> > > >> >> >>>>> > > > >> > >> >> if we >> > > >> >> >>>>> > > > >> > >> >> >>> > could >> > > >> >> >>>>> > > > >> > >> >> >>> > >> identify 1 or 2 >> strong >> > > use cases >> > > >> >> for >> > > >> >> >>>> headers >> > > >> >> >>>>> > in the >> > > >> >> >>>>> > > > space >> > > >> >> >>>>> > > > >> > for >> > > >> >> >>>>> > > > >> > >> >> >>> > third-party >> > > >> >> >>>>> > > > >> > >> >> >>> > >> vendors. For use >> cases >> > > within an >> > > >> >> >>>> organization, >> > > >> >> >>>>> > one >> > > >> >> >>>>> > > > could >> > > >> >> >>>>> > > > >> > always >> > > >> >> >>>>> > > > >> > >> >> use >> > > >> >> >>>>> > > > >> > >> >> >>> > other >> > > >> >> >>>>> > > > >> > >> >> >>> > >> approaches such as >> > > company-wise >> > > >> >> >>>> containers to >> > > >> >> >>>>> > get >> > > >> >> >>>>> > > > around >> > > >> >> >>>>> > > > >> w/o >> > > >> >> >>>>> > > > >> > >> >> >>> headers. I >> > > >> >> >>>>> > > > >> > >> >> >>> > >> went through the u= se >> > > cases in >> > > >> the >> > > >> >> KIP >> > > >> >> >>>> and in >> > > >> >> >>>>> > Radai's >> > > >> >> >>>>> > > > wiki >> > > >> >> >>>>> > > > >> ( >> > > >> >> >>>>> > > > >> > >> >> >>> > >> >> > > https://cwiki.apache.org/confl >> > > >> >> >>>>> > uence/display/KAFKA/A+ >> > > >> >> >>>>> > > > >> > >> >> >>> > Case+for+Kafka+Header= s >> > > >> >> >>>>> > > > >> > >> >> >>> > >> ). >> > > >> >> >>>>> > > > >> > >> >> >>> > >> The following are = the >> > > ones that >> > > >> >> that >> > > >> >> >> I >> > > >> >> >>>>> > understand and >> > > >> >> >>>>> > > > >> could >> > > >> >> >>>>> > > > >> > be >> > > >> >> >>>>> > > > >> > >> in >> > > >> >> >>>>> > > > >> > >> >> the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> third-party use ca= se >> > > category. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> A. content-type >> > > >> >> >>>>> > > > >> > >> >> >>> > >> It seems that in >> > general, >> > > >> >> >> content-type >> > > >> >> >>>> should >> > > >> >> >>>>> > be set >> > > >> >> >>>>> > > at >> > > >> >> >>>>> > > > >> the >> > > >> >> >>>>> > > > >> > >> topic >> > > >> >> >>>>> > > > >> > >> >> >>> level. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> Not sure if mixing >> > > messages with >> > > >> >> >>>> different >> > > >> >> >>>>> > content >> > > >> >> >>>>> > > > types >> > > >> >> >>>>> > > > >> > >> should be >> > > >> >> >>>>> > > > >> > >> >> >>> > >> encouraged. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> B. schema id >> > > >> >> >>>>> > > > >> > >> >> >>> > >> Since the value is >> > mostly >> > > >> useless >> > > >> >> >>>> without >> > > >> >> >>>>> > schema id, >> > > >> >> >>>>> > > it >> > > >> >> >>>>> > > > >> > seems >> > > >> >> >>>>> > > > >> > >> that >> > > >> >> >>>>> > > > >> > >> >> >>> > storing >> > > >> >> >>>>> > > > >> > >> >> >>> > >> the schema id >> together >> > > with >> > > >> >> >> serialized >> > > >> >> >>>> bytes >> > > >> >> >>>>> > in the >> > > >> >> >>>>> > > > value >> > > >> >> >>>>> > > > >> is >> > > >> >> >>>>> > > > >> > >> >> better? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> C. per message >> > encryption >> > > >> >> >>>>> > > > >> > >> >> >>> > >> One drawback of th= is >> > > approach is >> > > >> >> that >> > > >> >> >>>> this >> > > >> >> >>>>> > > > significantly >> > > >> >> >>>>> > > > >> > reduce >> > > >> >> >>>>> > > > >> > >> >> the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> effectiveness of >> > > compression, >> > > >> >> which >> > > >> >> >>>> happens on >> > > >> >> >>>>> > a set >> > > >> >> >>>>> > > of >> > > >> >> >>>>> > > > >> > >> serialized >> > > >> >> >>>>> > > > >> > >> >> >>> > >> messages. An >> > alternative >> > > is to >> > > >> >> enable >> > > >> >> >>>> SSL for >> > > >> >> >>>>> > wire >> > > >> >> >>>>> > > > >> > encryption >> > > >> >> >>>>> > > > >> > >> and >> > > >> >> >>>>> > > > >> > >> >> >>> rely >> > > >> >> >>>>> > > > >> > >> >> >>> > on >> > > >> >> >>>>> > > > >> > >> >> >>> > >> the storage system >> > (e.g. >> > > LUKS) >> > > >> >> for at >> > > >> >> >>>> rest >> > > >> >> >>>>> > > encryption. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> D. cluster ID for >> > > mirroring >> > > >> across >> > > >> >> >>>> Kafka >> > > >> >> >>>>> > clusters >> > > >> >> >>>>> > > > >> > >> >> >>> > >> This is actually >> > > interesting. >> > > >> >> Today, >> > > >> >> >>>> to avoid >> > > >> >> >>>>> > > > introducing >> > > >> >> >>>>> > > > >> > >> cycles >> > > >> >> >>>>> > > > >> > >> >> when >> > > >> >> >>>>> > > > >> > >> >> >>> > doing >> > > >> >> >>>>> > > > >> > >> >> >>> > >> mirroring across d= ata >> > > centers, >> > > >> one >> > > >> >> >>>> would >> > > >> >> >>>>> > either have >> > > >> >> >>>>> > > to >> > > >> >> >>>>> > > > >> set >> > > >> >> >>>>> > > > >> > up >> > > >> >> >>>>> > > > >> > >> two >> > > >> >> >>>>> > > > >> > >> >> >>> Kafka >> > > >> >> >>>>> > > > >> > >> >> >>> > >> clusters (a local = and >> > an >> > > >> >> aggregate) >> > > >> >> >>>> per data >> > > >> >> >>>>> > center >> > > >> >> >>>>> > > or >> > > >> >> >>>>> > > > >> > rename >> > > >> >> >>>>> > > > >> > >> >> topics. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> Neither is ideal. >> With >> > > headers, >> > > >> >> the >> > > >> >> >>>> producer >> > > >> >> >>>>> > could >> > > >> >> >>>>> > > tag >> > > >> >> >>>>> > > > >> each >> > > >> >> >>>>> > > > >> > >> >> message >> > > >> >> >>>>> > > > >> > >> >> >>> with >> > > >> >> >>>>> > > > >> > >> >> >>> > >> the producing clus= ter >> > ID >> > > in the >> > > >> >> >> header. >> > > >> >> >>>>> > MirrorMaker >> > > >> >> >>>>> > > > could >> > > >> >> >>>>> > > > >> > then >> > > >> >> >>>>> > > > >> > >> >> avoid >> > > >> >> >>>>> > > > >> > >> >> >>> > >> mirroring messages >> to a >> > > cluster >> > > >> if >> > > >> >> >>>> they are >> > > >> >> >>>>> > tagged >> > > >> >> >>>>> > > with >> > > >> >> >>>>> > > > >> the >> > > >> >> >>>>> > > > >> > >> same >> > > >> >> >>>>> > > > >> > >> >> >>> cluster >> > > >> >> >>>>> > > > >> > >> >> >>> > >> id. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> However, an >> alternative >> > > approach >> > > >> >> is >> > > >> >> >> to >> > > >> >> >>>>> > introduce sth >> > > >> >> >>>>> > > > like >> > > >> >> >>>>> > > > >> > >> >> >>> hierarchical >> > > >> >> >>>>> > > > >> > >> >> >>> > >> topic and store >> > messages >> > > from >> > > >> >> >> different >> > > >> >> >>>>> > clusters in >> > > >> >> >>>>> > > > >> > different >> > > >> >> >>>>> > > > >> > >> >> >>> partitions >> > > >> >> >>>>> > > > >> > >> >> >>> > >> under the same top= ic. >> > > This >> > > >> >> approach >> > > >> >> >>>> avoids >> > > >> >> >>>>> > filtering >> > > >> >> >>>>> > > > out >> > > >> >> >>>>> > > > >> > >> unneeded >> > > >> >> >>>>> > > > >> > >> >> >>> data >> > > >> >> >>>>> > > > >> > >> >> >>> > and >> > > >> >> >>>>> > > > >> > >> >> >>> > >> makes offset >> preserving >> > > easier >> > > >> to >> > > >> >> >>>> support. It >> > > >> >> >>>>> > may >> > > >> >> >>>>> > > make >> > > >> >> >>>>> > > > >> > >> compaction >> > > >> >> >>>>> > > > >> > >> >> >>> > trickier >> > > >> >> >>>>> > > > >> > >> >> >>> > >> though since the s= ame >> > > key may >> > > >> >> show up >> > > >> >> >>>> in >> > > >> >> >>>>> > different >> > > >> >> >>>>> > > > >> > partitions. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> E. record-level >> lineage >> > > >> >> >>>>> > > > >> > >> >> >>> > >> For example, a sou= rce >> > > connector >> > > >> >> could >> > > >> >> >>>> store in >> > > >> >> >>>>> > the >> > > >> >> >>>>> > > > message >> > > >> >> >>>>> > > > >> > the >> > > >> >> >>>>> > > > >> > >> >> >>> metadata >> > > >> >> >>>>> > > > >> > >> >> >>> > >> (e.g. UUID) of the >> > source >> > > >> record. >> > > >> >> >>>> Similarly, >> > > >> >> >>>>> > if a >> > > >> >> >>>>> > > > stream >> > > >> >> >>>>> > > > >> job >> > > >> >> >>>>> > > > >> > >> >> >>> transforms >> > > >> >> >>>>> > > > >> > >> >> >>> > >> messages from topi= c A >> > to >> > > topic >> > > >> B, >> > > >> >> the >> > > >> >> >>>> library >> > > >> >> >>>>> > could >> > > >> >> >>>>> > > > >> include >> > > >> >> >>>>> > > > >> > the >> > > >> >> >>>>> > > > >> > >> >> >>> source >> > > >> >> >>>>> > > > >> > >> >> >>> > >> message offset in >> each >> > > of the >> > > >> >> >>>> transformed >> > > >> >> >>>>> > message in >> > > >> >> >>>>> > > > the >> > > >> >> >>>>> > > > >> > >> header. >> > > >> >> >>>>> > > > >> > >> >> Not >> > > >> >> >>>>> > > > >> > >> >> >>> > sure >> > > >> >> >>> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> how widely useful >> > > record-level >> > > >> >> >> lineage >> > > >> >> >>>> is >> > > >> >> >>>>> > though >> > > >> >> >>>>> > > since >> > > >> >> >>>>> > > > the >> > > >> >> >>>>> > > > >> > >> >> overhead >> > > >> >> >>>>> > > > >> > >> >> >>> > could >> > > >> >> >>>>> > > > >> > >> >> >>> > >> be significant. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> F. auditing metada= ta >> > > >> >> >>>>> > > > >> > >> >> >>> > >> We could put thing= s >> > like >> > > >> >> >>>> clientId/host/user in >> > > >> >> >>>>> > the >> > > >> >> >>>>> > > > header >> > > >> >> >>>>> > > > >> in >> > > >> >> >>>>> > > > >> > >> each >> > > >> >> >>>>> > > > >> > >> >> >>> > message >> > > >> >> >>>>> > > > >> > >> >> >>> > >> for auditing. Thes= e >> > > metadata are >> > > >> >> >>>> really at the >> > > >> >> >>>>> > > producer >> > > >> >> >>>>> > > > >> > level >> > > >> >> >>>>> > > > >> > >> >> though. >> > > >> >> >>>>> > > > >> > >> >> >>> > So, a >> > > >> >> >>>>> > > > >> > >> >> >>> > >> more efficient way= is >> > to >> > > only >> > > >> >> >> include a >> > > >> >> >>>>> > "producerId" >> > > >> >> >>>>> > > > per >> > > >> >> >>>>> > > > >> > >> message >> > > >> >> >>>>> > > > >> > >> >> and >> > > >> >> >>>>> > > > >> > >> >> >>> > send >> > > >> >> >>>>> > > > >> > >> >> >>> > >> the producerId -> >> > > metadata >> > > >> mapping >> > > >> >> >>>>> > independently. >> > > >> >> >>>>> > > > KIP-98 >> > > >> >> >>>>> > > > >> is >> > > >> >> >>>>> > > > >> > >> >> actually >> > > >> >> >>>>> > > > >> > >> >> >>> > >> proposing includin= g >> > such >> > > a >> > > >> >> producerId >> > > >> >> >>>> natively >> > > >> >> >>>>> > in the >> > > >> >> >>>>> > > > >> > message. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> So, overall, I not >> sure >> > > that I >> > > >> am >> > > >> >> >> fully >> > > >> >> >>>>> > convinced of >> > > >> >> >>>>> > > > the >> > > >> >> >>>>> > > > >> > strong >> > > >> >> >>>>> > > > >> > >> >> >>> > third-party >> > > >> >> >>>>> > > > >> > >> >> >>> > >> use cases of heade= rs >> > yet. >> > > >> Perhaps >> > > >> >> we >> > > >> >> >>>> could >> > > >> >> >>>>> > discuss a >> > > >> >> >>>>> > > > bit >> > > >> >> >>>>> > > > >> > more >> > > >> >> >>>>> > > > >> > >> to >> > > >> >> >>>>> > > > >> > >> >> make >> > > >> >> >>>>> > > > >> > >> >> >>> > one >> > > >> >> >>>>> > > > >> > >> >> >>> > >> or two really >> > convincing >> > > use >> > > >> >> cases. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> Another orthogonal >> > > question is >> > > >> >> >>>> whether header >> > > >> >> >>>>> > should >> > > >> >> >>>>> > > > be >> > > >> >> >>>>> > > > >> > >> exposed >> > > >> >> >>>>> > > > >> > >> >> in >> > > >> >> >>>>> > > > >> > >> >> >>> > stream >> > > >> >> >>>>> > > > >> > >> >> >>> > >> processing systems >> such >> > > Kafka >> > > >> >> stream, >> > > >> >> >>>> Samza, >> > > >> >> >>>>> > and >> > > >> >> >>>>> > > Spark >> > > >> >> >>>>> > > > >> > >> streaming. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> Currently, those >> > systems >> > > just >> > > >> deal >> > > >> >> >> with >> > > >> >> >>>>> > key/value >> > > >> >> >>>>> > > > pairs. >> > > >> >> >>>>> > > > >> > >> Should we >> > > >> >> >>>>> > > > >> > >> >> >>> > expose a >> > > >> >> >>>>> > > > >> > >> >> >>> > >> third thing header >> > there >> > > too or >> > > >> >> >>>> somehow map >> > > >> >> >>>>> > header to >> > > >> >> >>>>> > > > key >> > > >> >> >>>>> > > > >> or >> > > >> >> >>>>> > > > >> > >> >> value? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> Thanks, >> > > >> >> >>>>> > > > >> > >> >> >>> > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> Jun >> > > >> >> >>>>> > > > >> > >> >> >>> > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> On Tue, Nov 29, 20= 16 >> at >> > > 3:35 AM, >> > > >> >> >>>> Michael >> > > >> >> >>>>> > Pearce < >> > > >> >> >>>>> > > > >> > >> >> >>> Michael.Pearce@ig.com> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> wrote: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > I assume, that >> after >> > a >> > > period >> > > >> >> of a >> > > >> >> >>>> week, >> > > >> >> >>>>> > that there >> > > >> >> >>>>> > > > is >> > > >> >> >>>>> > > > >> no >> > > >> >> >>>>> > > > >> > >> >> concerns >> > > >> >> >>>>> > > > >> > >> >> >>> now >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > with points 1, a= nd >> 2 >> > > and now >> > > >> we >> > > >> >> >> have >> > > >> >> >>>>> > agreement that >> > > >> >> >>>>> > > > >> > headers >> > > >> >> >>>>> > > > >> > >> are >> > > >> >> >>>>> > > > >> > >> >> >>> useful >> > > >> >> >>>>> > > > >> > >> >> >>> > >> and >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > needed in Kafka.= As >> > > such if >> > > >> put >> > > >> >> to >> > > >> >> >> a >> > > >> >> >>>> KIP >> > > >> >> >>>>> > vote, this >> > > >> >> >>>>> > > > >> > wouldn=E2=80=99t >> > > >> >> >>>>> > > > >> > >> be >> > > >> >> >>>>> > > > >> > >> >> a >> > > >> >> >>>>> > > > >> > >> >> >>> > reason >> > > >> >> >>>>> > > > >> > >> >> >>> > >> to >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > reject. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > @ >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > Ignacio on point >> 4). >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > I think for purp= ose >> > of >> > > getting >> > > >> >> this >> > > >> >> >>>> KIP >> > > >> >> >>>>> > moving past >> > > >> >> >>>>> > > > >> this, >> > > >> >> >>>>> > > > >> > we >> > > >> >> >>>>> > > > >> > >> can >> > > >> >> >>>>> > > > >> > >> >> >>> state >> > > >> >> >>>>> > > > >> > >> >> >>> > >> the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > key will be a 4 >> bytes >> > > space >> > > >> that >> > > >> >> >> can >> > > >> >> >>>> will be >> > > >> >> >>>>> > > > naturally >> > > >> >> >>>>> > > > >> > >> >> interpreted >> > > >> >> >>>>> > > > >> > >> >> >>> as >> > > >> >> >>>>> > > > >> > >> >> >>> > an >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > Int32 (if >> namespacing >> > > is later >> > > >> >> >>>> wanted you can >> > > >> >> >>>>> > > easily >> > > >> >> >>>>> > > > >> split >> > > >> >> >>>>> > > > >> > >> this >> > > >> >> >>>>> > > > >> > >> >> >>> into >> > > >> >> >>>>> > > > >> > >> >> >>> > two >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > int16 spaces), f= rom >> > > the wire >> > > >> >> >> protocol >> > > >> >> >>>>> > > implementation >> > > >> >> >>>>> > > > >> this >> > > >> >> >>>>> > > > >> > >> makes >> > > >> >> >>>>> > > > >> > >> >> no >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > difference I don= =E2=80=99t >> > > believe. Is >> > > >> >> this >> > > >> >> >>>>> > reasonable to >> > > >> >> >>>>> > > > all? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > On 5) as per poi= nt >> 4 >> > > therefor >> > > >> >> happy >> > > >> >> >>>> we keep >> > > >> >> >>>>> > with 32 >> > > >> >> >>>>> > > > >> bits. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > On 18/11/2016, >> > 20:34, " >> > > >> >> >>>>> > ignacio.solis@gmail.com on >> > > >> >> >>>>> > > > >> behalf >> > > >> >> >>>>> > > > >> > of >> > > >> >> >>>>> > > > >> > >> >> >>> Ignacio >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > Solis" < >> > > >> ignacio.solis@gmail.com >> > > >> >> on >> > > >> >> >>>> behalf of >> > > >> >> >>>>> > > > >> > isolis@igso.net >> > > >> >> >>>>> > > > >> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> wrote: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > Summary: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > 3) Yes - Hea= der >> > > value as >> > > >> >> byte[] >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > 4a) Int,Int = - >> No >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > 4b) Int - Ye= s >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > 4c) String - >> > > Reluctant >> > > >> maybe >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > 5) I believe >> the >> > > header >> > > >> >> system >> > > >> >> >>>> should >> > > >> >> >>>>> > take a >> > > >> >> >>>>> > > > single >> > > >> >> >>>>> > > > >> > >> int. I >> > > >> >> >>>>> > > > >> > >> >> >>> think >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > 32bits is >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > a good size,= if >> > > you want >> > > >> to >> > > >> >> >>>> interpret >> > > >> >> >>>>> > this as >> > > >> >> >>>>> > > to >> > > >> >> >>>>> > > > >> 16bit >> > > >> >> >>>>> > > > >> > >> >> numbers >> > > >> >> >>>>> > > > >> > >> >> >>> in >> > > >> >> >>>>> > > > >> > >> >> >>> > the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > layer >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > above go rig= ht >> > > ahead. If >> > > >> >> >>>> somebody wants >> > > >> >> >>>>> > to >> > > >> >> >>>>> > > argue >> > > >> >> >>>>> > > > >> for >> > > >> >> >>>>> > > > >> > 16 >> > > >> >> >>>>> > > > >> > >> >> bits >> > > >> >> >>>>> > > > >> > >> >> >>> or >> > > >> >> >>>>> > > > >> > >> >> >>> > 64 >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > bits of >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > header key >> space >> > I >> > > would >> > > >> >> >> listen. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > Discussion: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > Dividing the >> key >> > > space >> > > >> into >> > > >> >> >>>> sub_key_1 and >> > > >> >> >>>>> > > > sub_key_2 >> > > >> >> >>>>> > > > >> > >> makes no >> > > >> >> >>>>> > > > >> > >> >> >>> > sense to >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > me at >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > this layer. >> Are >> > > we going >> > > >> to >> > > >> >> >>>> start >> > > >> >> >>>>> > providing >> > > >> >> >>>>> > > > APIs to >> > > >> >> >>>>> > > > >> > get >> > > >> >> >>>>> > > > >> > >> all >> > > >> >> >>>>> > > > >> > >> >> >>> the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > sub_key_1s? = or >> > all >> > > the >> > > >> >> >>>> sub_key_2s? If >> > > >> >> >>>>> > there is >> > > >> >> >>>>> > > > no >> > > >> >> >>>>> > > > >> > >> >> >>> distinguishing >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > functions >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > that are >> applied >> > > to each >> > > >> one >> > > >> >> >>>> then they >> > > >> >> >>>>> > should >> > > >> >> >>>>> > > be >> > > >> >> >>>>> > > > a >> > > >> >> >>>>> > > > >> > single >> > > >> >> >>>>> > > > >> > >> >> >>> value. >> > > >> >> >>>>> > > > >> > >> >> >>> > At >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > this >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > layer all we= 're >> > > doing is >> > > >> >> >>>> equality. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > If the above >> > layer >> > > wants >> > > >> to >> > > >> >> >>>> interpret >> > > >> >> >>>>> > this as >> > > >> >> >>>>> > > 2, >> > > >> >> >>>>> > > > 3 >> > > >> >> >>>>> > > > >> or >> > > >> >> >>>>> > > > >> > >> more >> > > >> >> >>>>> > > > >> > >> >> >>> values >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > that's a >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > different >> > > question. I >> > > >> >> >>>> personally think >> > > >> >> >>>>> > it's >> > > >> >> >>>>> > > all >> > > >> >> >>>>> > > > one >> > > >> >> >>>>> > > > >> > >> >> keyspace >> > > >> >> >>>>> > > > >> > >> >> >>> > that is >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > getting >> assigned >> > > using >> > > >> some >> > > >> >> >>>> structure, >> > > >> >> >>>>> > but if >> > > >> >> >>>>> > > you >> > > >> >> >>>>> > > > >> > want to >> > > >> >> >>>>> > > > >> > >> >> >>> > sub-assign >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > parts >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > of it then >> that's >> > > fine. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > The same >> > discussion >> > > >> applies >> > > >> >> to >> > > >> >> >>>> strings. >> > > >> >> >>>>> > If >> > > >> >> >>>>> > > > somebody >> > > >> >> >>>>> > > > >> > >> argued >> > > >> >> >>>>> > > > >> > >> >> for >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > strings, >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > would we be >> > > arguing to >> > > >> >> divide >> > > >> >> >> the >> > > >> >> >>>>> > strings with >> > > >> >> >>>>> > > > dots >> > > >> >> >>>>> > > > >> > ('.') >> > > >> >> >>>>> > > > >> > >> >> as a >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > requirement? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > Would we wan= t >> > them >> > > to give >> > > >> >> us >> > > >> >> >> the >> > > >> >> >>>>> > different >> > > >> >> >>>>> > > name >> > > >> >> >>>>> > > > >> > segments >> > > >> >> >>>>> > > > >> > >> >> >>> > separately? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > Would we be >> > > performing any >> > > >> >> >>>> actions on >> > > >> >> >>>>> > this key >> > > >> >> >>>>> > > > other >> > > >> >> >>>>> > > > >> > than >> > > >> >> >>>>> > > > >> > >> >> >>> > matching? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > Nacho >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > On Fri, Nov = 18, >> > > 2016 at >> > > >> 9:30 >> > > >> >> >> AM, >> > > >> >> >>>> Michael >> > > >> >> >>>>> > > Pearce < >> > > >> >> >>>>> > > > >> > >> >> >>> > >> >> Michael.Pearce@ig.com >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > wrote: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > #jay #jun = any >> > > concerns >> > > >> on >> > > >> >> 1 >> > > >> >> >>>> and 2 >> > > >> >> >>>>> > still? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > @all >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > To get thi= s >> > > moving >> > > >> along a >> > > >> >> >> bit >> > > >> >> >>>> more >> > > >> >> >>>>> > I'd also >> > > >> >> >>>>> > > > like >> > > >> >> >>>>> > > > >> to >> > > >> >> >>>>> > > > >> > >> ask >> > > >> >> >>>>> > > > >> > >> >> to >> > > >> >> >>>>> > > > >> > >> >> >>> get >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > clarity on >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > the below >> last >> > > points: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > 3) I belie= ve >> > > we're all >> > > >> >> >> roughly >> > > >> >> >>>> happy >> > > >> >> >>>>> > with the >> > > >> >> >>>>> > > > >> header >> > > >> >> >>>>> > > > >> > >> value >> > > >> >> >>>>> > > > >> > >> >> >>> > being a >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > byte[]? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > 4) I belie= ve >> > > consensus >> > > >> has >> > > >> >> >>>> been for an >> > > >> >> >>>>> > > > namespace >> > > >> >> >>>>> > > > >> > based >> > > >> >> >>>>> > > > >> > >> int >> > > >> >> >>>>> > > > >> > >> >> >>> > approach >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > {int,int} = for >> > > the key. >> > > >> Any >> > > >> >> >>>> objections >> > > >> >> >>>>> > if this >> > > >> >> >>>>> > > > is >> > > >> >> >>>>> > > > >> > what >> > > >> >> >>>>> > > > >> > >> we >> > > >> >> >>>>> > > > >> > >> >> go >> > > >> >> >>>>> > > > >> > >> >> >>> > with? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > 5) as we h= ave >> > if >> > > >> >> assumption >> > > >> >> >> in >> > > >> >> >>>> (4) is >> > > >> >> >>>>> > > correct, >> > > >> >> >>>>> > > > >> > >> {int,int} >> > > >> >> >>>>> > > > >> > >> >> >>> keys. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > Should bot= h >> > > int's be >> > > >> >> int16 or >> > > >> >> >>>> int32? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > I'm for th= em >> > > being >> > > >> int16(2 >> > > >> >> >>>> bytes) as >> > > >> >> >>>>> > combined >> > > >> >> >>>>> > > > is >> > > >> >> >>>>> > > > >> > space >> > > >> >> >>>>> > > > >> > >> of >> > > >> >> >>>>> > > > >> > >> >> >>> > 4bytes as >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > per >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > original a= nd >> > > gives >> > > >> plenty >> > > >> >> of >> > > >> >> >>>>> > combinations for >> > > >> >> >>>>> > > > the >> > > >> >> >>>>> > > > >> > >> >> >>> foreseeable, >> > > >> >> >>>>> > > > >> > >> >> >>> > and >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > keeps >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > the overhe= ad >> > > small. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > Do we see = any >> > > benefit in >> > > >> >> >>>> another kip >> > > >> >> >>>>> > call to >> > > >> >> >>>>> > > > >> discuss >> > > >> >> >>>>> > > > >> > >> >> these at >> > > >> >> >>>>> > > > >> > >> >> >>> > all? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > Cheers >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > Mike >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> >> >> ______________________________ >> > > >> >> >>>>> > __________ >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > From: K >> > Burstev < >> > > >> >> >>>> k.burstev@yandex.com> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > Sent: Frid= ay, >> > > November >> > > >> 18, >> > > >> >> >> 2016 >> > > >> >> >>>>> > 7:07:07 AM >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > To: >> > > >> dev@kafka.apache.org >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > Subject: R= e: >> > > [DISCUSS] >> > > >> >> KIP-82 >> > > >> >> >>>> - Add >> > > >> >> >>>>> > Record >> > > >> >> >>>>> > > > Headers >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > For what i= t >> is >> > > worth >> > > >> also >> > > >> >> i >> > > >> >> >>>> agree. As >> > > >> >> >>>>> > a user: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > 1) Yes - >> > > Headers are >> > > >> >> >>>> worthwhile >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > 2) Yes - >> > > Headers should >> > > >> >> be a >> > > >> >> >>>> top level >> > > >> >> >>>>> > > option >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > 14.11.2016= , >> > > 21:15, >> > > >> >> "Ignacio >> > > >> >> >>>> Solis" < >> > > >> >> >>>>> > > > >> isolis@igso.net >> > > >> >> >>>>> > > > >> > >: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > > 1) Yes - >> > > Headers are >> > > >> >> >>>> worthwhile >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > > 2) Yes - >> > > Headers >> > > >> should >> > > >> >> be >> > > >> >> >> a >> > > >> >> >>>> top >> > > >> >> >>>>> > level >> > > >> >> >>>>> > > option >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > > On Mon, = Nov >> > > 14, 2016 >> > > >> at >> > > >> >> >> 9:16 >> > > >> >> >>>> AM, >> > > >> >> >>>>> > Michael >> > > >> >> >>>>> > > > Pearce >> > > >> >> >>>>> > > > >> < >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > Michael.Pearce@ig.com> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > > wrote: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> Hi Rog= er, >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> The ki= p >> > > >> >> details/examples >> > > >> >> >>>> the >> > > >> >> >>>>> > original >> > > >> >> >>>>> > > > proposal >> > > >> >> >>>>> > > > >> > for >> > > >> >> >>>>> > > > >> > >> key >> > > >> >> >>>>> > > > >> > >> >> >>> > spacing >> > > >> >> >>>>> > > > >> > >> >> >>> > >> , >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > not >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> new >> > > mentioned as per >> > > >> >> >>>> discussion >> > > >> >> >>>>> > namespace >> > > >> >> >>>>> > > > >> idea. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> We wil= l >> > need >> > > to >> > > >> update >> > > >> >> >> the >> > > >> >> >>>> kip, >> > > >> >> >>>>> > when we >> > > >> >> >>>>> > > get >> > > >> >> >>>>> > > > >> > >> agreement >> > > >> >> >>>>> > > > >> > >> >> >>> this >> > > >> >> >>>>> > > > >> > >> >> >>> > is a >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > better >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> approa= ch >> > > (which >> > > >> seems >> > > >> >> to >> > > >> >> >>>> be the >> > > >> >> >>>>> > case if I >> > > >> >> >>>>> > > > have >> > > >> >> >>>>> > > > >> > >> >> understood >> > > >> >> >>>>> > > > >> > >> >> >>> > the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > general >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> feelin= g >> in >> > > the >> > > >> >> >>>> conversation) >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> Re the >> > > variable >> > > >> ints, >> > > >> >> at >> > > >> >> >>>> very >> > > >> >> >>>>> > early stage >> > > >> >> >>>>> > > > we >> > > >> >> >>>>> > > > >> did >> > > >> >> >>>>> > > > >> > >> think >> > > >> >> >>>>> > > > >> > >> >> >>> about >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > this. I >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > think >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> the ad= ded >> > > complexity >> > > >> >> for >> > > >> >> >>>> the >> > > >> >> >>>>> > saving isn't >> > > >> >> >>>>> > > > >> worth >> > > >> >> >>>>> > > > >> > it. >> > > >> >> >>>>> > > > >> > >> >> I'd >> > > >> >> >>>>> > > > >> > >> >> >>> > rather >> > > >> >> >>>>> > > > >> > >> >> >>> > >> go >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > with, if >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> we wan= t >> to >> > > reduce >> > > >> >> >>>> overheads and >> > > >> >> >>>>> > size >> > > >> >> >>>>> > > int16 >> > > >> >> >>>>> > > > >> > (2bytes) >> > > >> >> >>>>> > > > >> > >> >> keys >> > > >> >> >>>>> > > > >> > >> >> >>> as >> > > >> >> >>>>> > > > >> > >> >> >>> > it >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > keeps it >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> simple= . >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> On the >> note >> > > of no >> > > >> >> >> headers, >> > > >> >> >>>> there >> > > >> >> >>>>> > is as >> > > >> >> >>>>> > > per >> > > >> >> >>>>> > > > the >> > > >> >> >>>>> > > > >> > kip >> > > >> >> >>>>> > > > >> > >> as >> > > >> >> >>>>> > > > >> > >> >> we >> > > >> >> >>>>> > > > >> > >> >> >>> > use an >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > attribute >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> bit to >> > > denote if >> > > >> >> headers >> > > >> >> >>>> are >> > > >> >> >>>>> > present or >> > > >> >> >>>>> > > > not as >> > > >> >> >>>>> > > > >> > such >> > > >> >> >>>>> > > > >> > >> >> >>> > provides a >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > zero >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> overhe= ad >> > > currently >> > > >> if >> > > >> >> >>>> headers are >> > > >> >> >>>>> > not >> > > >> >> >>>>> > > used. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> I thin= k >> as >> > > radai >> > > >> >> mentions >> > > >> >> >>>> would be >> > > >> >> >>>>> > good >> > > >> >> >>>>> > > > first >> > > >> >> >>>>> > > > >> > if we >> > > >> >> >>>>> > > > >> > >> >> can >> > > >> >> >>>>> > > > >> > >> >> >>> get >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > clarity if >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > do >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> we now >> have >> > > general >> > > >> >> >>>> consensus that >> > > >> >> >>>>> > (1) >> > > >> >> >>>>> > > > headers >> > > >> >> >>>>> > > > >> > are >> > > >> >> >>>>> > > > >> > >> >> >>> > worthwhile >> > > >> >> >>>>> > > > >> > >> >> >>> > >> and >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > useful, >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> and (2= ) >> we >> > > want it >> > > >> as >> > > >> >> a >> > > >> >> >>>> top level >> > > >> >> >>>>> > entity. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> Just t= o >> > > state the >> > > >> >> >> obvious i >> > > >> >> >>>>> > believe (1) >> > > >> >> >>>>> > > > >> headers >> > > >> >> >>>>> > > > >> > are >> > > >> >> >>>>> > > > >> > >> >> >>> > worthwhile >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > and (2) >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> agree = as >> a >> > > top level >> > > >> >> >>>> entity. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> Cheers >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> Mike >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> > > >> >> >>>> ______________________________ >> > > >> >> >>>>> > __________ >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> From: >> Roger >> > > Hoover < >> > > >> >> >>>>> > > roger.hoover@gmail.com >> > > >> >> >>>>> > > > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> Sent: >> > > Wednesday, >> > > >> >> November >> > > >> >> >>>> 9, 2016 >> > > >> >> >>>>> > 9:10:47 >> > > >> >> >>>>> > > > PM >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> To: >> > > >> >> dev@kafka.apache.org >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> Subjec= t: >> > Re: >> > > >> [DISCUSS] >> > > >> >> >>>> KIP-82 - Add >> > > >> >> >>>>> > > Record >> > > >> >> >>>>> > > > >> > Headers >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> Sorry = for >> > > going a >> > > >> >> little >> > > >> >> >>>> in the >> > > >> >> >>>>> > weeds but >> > > >> >> >>>>> > > > >> thanks >> > > >> >> >>>>> > > > >> > >> for >> > > >> >> >>>>> > > > >> > >> >> the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> replies >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > regarding >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> varint= . >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> Agreed >> that >> > > a prefix >> > > >> >> and >> > > >> >> >>>> {int, >> > > >> >> >>>>> > int} can >> > > >> >> >>>>> > > be >> > > >> >> >>>>> > > > the >> > > >> >> >>>>> > > > >> > >> same. >> > > >> >> >>>>> > > > >> > >> >> It >> > > >> >> >>>>> > > > >> > >> >> >>> > doesn't >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > look >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > like >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> that's >> what >> > > the KIP >> > > >> is >> > > >> >> >>>> saying the >> > > >> >> >>>>> > "Open" >> > > >> >> >>>>> > > > >> > section. >> > > >> >> >>>>> > > > >> > >> The >> > > >> >> >>>>> > > > >> > >> >> >>> > example >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > shows >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> 210000= 1 >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> for Ne= w >> > > Relic and >> > > >> >> 210002 >> > > >> >> >>>> for App >> > > >> >> >>>>> > Dynamics >> > > >> >> >>>>> > > > >> > implying >> > > >> >> >>>>> > > > >> > >> >> that >> > > >> >> >>>>> > > > >> > >> >> >>> the >> > > >> >> >>>>> > > > >> > >> >> >>> > New >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > Relic >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> > organization >> > > will >> > > >> have >> > > >> >> >>>> only a >> > > >> >> >>>>> > single >> > > >> >> >>>>> > > > header id >> > > >> >> >>>>> > > > >> > to >> > > >> >> >>>>> > > > >> > >> work >> > > >> >> >>>>> > > > >> > >> >> >>> > with. Or >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > is >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > 2100001 >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> a pref= ix? >> > > The main >> > > >> >> point >> > > >> >> >>>> of a >> > > >> >> >>>>> > namespace >> > > >> >> >>>>> > > or >> > > >> >> >>>>> > > > >> > prefix >> > > >> >> >>>>> > > > >> > >> is >> > > >> >> >>>>> > > > >> > >> >> to >> > > >> >> >>>>> > > > >> > >> >> >>> > reduce >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> overhe= ad >> of >> > > config >> > > >> >> >> mapping >> > > >> >> >>>> or >> > > >> >> >>>>> > > registration >> > > >> >> >>>>> > > > >> > >> depending >> > > >> >> >>>>> > > > >> > >> >> on >> > > >> >> >>>>> > > > >> > >> >> >>> how >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> > > namespaces/prefixes >> > > >> >> are >> > > >> >> >>>> managed. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> Would >> love >> > > to hear >> > > >> >> more >> > > >> >> >>>> feedback >> > > >> >> >>>>> > on the >> > > >> >> >>>>> > > > >> > >> higher-level >> > > >> >> >>>>> > > > >> > >> >> >>> > questions >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > though... >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> Cheers= , >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> Roger >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> On Wed= , >> Nov >> > > 9, 2016 >> > > >> at >> > > >> >> >>>> 11:38 AM, >> > > >> >> >>>>> > radai < >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > radai.rosenblatt@gmail.com> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > wrote: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > I th= ink >> > > this >> > > >> >> discussion >> > > >> >> >>>> is >> > > >> >> >>>>> > getting a >> > > >> >> >>>>> > > bit >> > > >> >> >>>>> > > > >> into >> > > >> >> >>>>> > > > >> > the >> > > >> >> >>>>> > > > >> > >> >> >>> weeds on >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > technical >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> > > implementation >> > > >> >> details. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > I'd >> liek >> > > to step >> > > >> >> back a >> > > >> >> >>>> minute >> > > >> >> >>>>> > and try >> > > >> >> >>>>> > > > and >> > > >> >> >>>>> > > > >> > >> establish >> > > >> >> >>>>> > > > >> > >> >> >>> > where we >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > are in >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > larg= er >> > > picture: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> > (re-wording >> > > >> nacho's >> > > >> >> >> last >> > > >> >> >>>>> > paragraph) >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > 1. a= re >> we >> > > all in >> > > >> >> >>>> agreement that >> > > >> >> >>>>> > headers >> > > >> >> >>>>> > > > are >> > > >> >> >>>>> > > > >> a >> > > >> >> >>>>> > > > >> > >> >> >>> worthwhile >> > > >> >> >>>>> > > > >> > >> >> >>> > and >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > useful >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> addition >> > > to have? >> > > >> >> this >> > > >> >> >>>> was >> > > >> >> >>>>> > contested >> > > >> >> >>>>> > > > early >> > > >> >> >>>>> > > > >> on >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > 2. a= re >> we >> > > all in >> > > >> >> >>>> agreement on >> > > >> >> >>>>> > headers >> > > >> >> >>>>> > > as >> > > >> >> >>>>> > > > top >> > > >> >> >>>>> > > > >> > >> level >> > > >> >> >>>>> > > > >> > >> >> >>> entity >> > > >> >> >>>>> > > > >> > >> >> >>> > vs >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > headers >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> > > squirreled-away in >> > > >> >> V? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > if >> there >> > > are still >> > > >> >> >>>> concerns >> > > >> >> >>>>> > around >> > > >> >> >>>>> > > these >> > > >> >> >>>>> > > > #2 >> > > >> >> >>>>> > > > >> > >> points >> > > >> >> >>>>> > > > >> > >> >> >>> (#jay? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > #jun?)? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > (and >> now >> > > back to >> > > >> our >> > > >> >> >>>> normal >> > > >> >> >>>>> > programming >> > > >> >> >>>>> > > > ...) >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > vari= nts >> > > are nice. >> > > >> >> >> having >> > > >> >> >>>> said >> > > >> >> >>>>> > that, its >> > > >> >> >>>>> > > > >> adding >> > > >> >> >>>>> > > > >> > >> >> >>> complexity >> > > >> >> >>>>> > > > >> > >> >> >>> > >> (see >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> > > >> >> >>>> https://github.com/addthis/ >> > > >> >> >>>>> > > > >> > >> >> stream-lib/blob/master/src/ >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> > > >> >> >>>> main/java/com/clearspring/ >> > > >> >> >>>>> > > > >> > >> >> analytics/util/Varint.java >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > as 1= st >> > > google >> > > >> >> result) >> > > >> >> >>>> and would >> > > >> >> >>>>> > require >> > > >> >> >>>>> > > > >> anyone >> > > >> >> >>>>> > > > >> > >> >> writing >> > > >> >> >>>>> > > > >> > >> >> >>> > other >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > clients >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > (C? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > Pyth= on? >> > > Go? Bash? >> > > >> >> ;-) ) >> > > >> >> >>>> to >> > > >> >> >>>>> > > get/implement >> > > >> >> >>>>> > > > the >> > > >> >> >>>>> > > > >> > >> same, >> > > >> >> >>>>> > > > >> > >> >> and >> > > >> >> >>>>> > > > >> > >> >> >>> for >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > relatively >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > litt= le >> > > gain (int >> > > >> vs >> > > >> >> >>>> string is >> > > >> >> >>>>> > order of >> > > >> >> >>>>> > > > >> > magnitude, >> > > >> >> >>>>> > > > >> > >> >> this >> > > >> >> >>>>> > > > >> > >> >> >>> > isnt). >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > int >> > > namespacing vs >> > > >> >> >> {int, >> > > >> >> >>>> int} >> > > >> >> >>>>> > > namespacing >> > > >> >> >>>>> > > > >> are >> > > >> >> >>>>> > > > >> > >> >> basically >> > > >> >> >>>>> > > > >> > >> >> >>> > the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > same >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > thing - >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > your= e >> > just >> > > >> >> namespacing >> > > >> >> >>>> an int64 >> > > >> >> >>>>> > and >> > > >> >> >>>>> > > > giving >> > > >> >> >>>>> > > > >> > people >> > > >> >> >>>>> > > > >> > >> >> while >> > > >> >> >>>>> > > > >> > >> >> >>> > 2^32 >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > ranges >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > at a >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > time= . >> the >> > > part i >> > > >> >> like >> > > >> >> >>>> about this >> > > >> >> >>>>> > is >> > > >> >> >>>>> > > > letting >> > > >> >> >>>>> > > > >> > >> people >> > > >> >> >>>>> > > > >> > >> >> >>> have a >> > > >> >> >>>>> > > > >> > >> >> >>> > >> large >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > swath of >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > numb= ers >> > > with one >> > > >> >> >>>> registration so >> > > >> >> >>>>> > they >> > > >> >> >>>>> > > > dont >> > > >> >> >>>>> > > > >> > have >> > > >> >> >>>>> > > > >> > >> to >> > > >> >> >>>>> > > > >> > >> >> come >> > > >> >> >>>>> > > > >> > >> >> >>> > back >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > for >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > every >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > sing= le >> > > >> plugin/header >> > > >> >> >>>> they want to >> > > >> >> >>>>> > > > "reserve". >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > On W= ed, >> > > Nov 9, >> > > >> 2016 >> > > >> >> at >> > > >> >> >>>> 11:01 AM, >> > > >> >> >>>>> > Roger >> > > >> >> >>>>> > > > >> Hoover >> > > >> >> >>>>> > > > >> > < >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > roger.hoover@gmail.com> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > wrot= e: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > Si= nce >> > > some of >> > > >> the >> > > >> >> >>>> debate has >> > > >> >> >>>>> > been >> > > >> >> >>>>> > > about >> > > >> >> >>>>> > > > >> > >> overhead + >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > performance, I'm >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> > > wondering if we >> > > >> >> have >> > > >> >> >>>>> > considered a >> > > >> >> >>>>> > > > varint >> > > >> >> >>>>> > > > >> > >> encoding >> > > >> >> >>>>> > > > >> > >> >> ( >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> > > >> >> >>>> https://developers.google.com/ >> > > >> >> >>>>> > > > >> > >> >> protocol-buffers/docs/ >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > encoding#varints) >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > for >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > th= e >> > > header >> > > >> length >> > > >> >> >>>> field (int32 >> > > >> >> >>>>> > in the >> > > >> >> >>>>> > > > >> > proposal) >> > > >> >> >>>>> > > > >> > >> >> and >> > > >> >> >>>>> > > > >> > >> >> >>> for >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > header >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > ids? If >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > you >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > do= n't >> > use >> > > >> headers, >> > > >> >> >> the >> > > >> >> >>>>> > overhead would >> > > >> >> >>>>> > > > be a >> > > >> >> >>>>> > > > >> > >> single >> > > >> >> >>>>> > > > >> > >> >> >>> byte >> > > >> >> >>>>> > > > >> > >> >> >>> > and >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > for each >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > head= er >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > id= < >> > 128 >> > > would >> > > >> >> also >> > > >> >> >>>> need only a >> > > >> >> >>>>> > > single >> > > >> >> >>>>> > > > >> byte? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > On >> Wed, >> > > Nov 9, >> > > >> >> 2016 >> > > >> >> >> at >> > > >> >> >>>> 6:43 AM, >> > > >> >> >>>>> > > radai < >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > >> > > radai.rosenblatt@gmail.com> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > wrot= e: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> > > @magnus - and >> > > >> >> very >> > > >> >> >>>> dangerous >> > > >> >> >>>>> > (youre >> > > >> >> >>>>> > > > >> > >> essentially >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > downloading and >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> executing >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> > > arbitrary code >> > > >> >> off >> > > >> >> >>>> the >> > > >> >> >>>>> > internet on >> > > >> >> >>>>> > > > your >> > > >> >> >>>>> > > > >> > >> servers >> > > >> >> >>>>> > > > >> > >> >> ... >> > > >> >> >>>>> > > > >> > >> >> >>> > bad >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > idea >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > without >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> a >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> > > sandbox, even >> > > >> >> with) >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = as >> > for >> > > it >> > > >> being >> > > >> >> a >> > > >> >> >>>> purely >> > > >> >> >>>>> > > > administrative >> > > >> >> >>>>> > > > >> > task >> > > >> >> >>>>> > > > >> > >> - i >> > > >> >> >>>>> > > > >> > >> >> >>> > >> disagree. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = i >> > wish >> > > it >> > > >> would, >> > > >> >> >>>> really, >> > > >> >> >>>>> > because >> > > >> >> >>>>> > > > then my >> > > >> >> >>>>> > > > >> > >> earlier >> > > >> >> >>>>> > > > >> > >> >> >>> > point on >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> > complexity >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > of >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = the >> > > remapping >> > > >> >> >>>> process would >> > > >> >> >>>>> > be >> > > >> >> >>>>> > > > invalid, >> > > >> >> >>>>> > > > >> > but >> > > >> >> >>>>> > > > >> > >> at >> > > >> >> >>>>> > > > >> > >> >> >>> > linkedin, >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > for >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > example, >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > we >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> (the >> > > team im >> > > >> in) >> > > >> >> >> run >> > > >> >> >>>> kafka >> > > >> >> >>>>> > as a >> > > >> >> >>>>> > > > service. >> > > >> >> >>>>> > > > >> > we >> > > >> >> >>>>> > > > >> > >> dont >> > > >> >> >>>>> > > > >> > >> >> >>> > really >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > know >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > what our >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > us= ers >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> > > (developing >> > > >> >> >>>> applications >> > > >> >> >>>>> > that use >> > > >> >> >>>>> > > > kafka) >> > > >> >> >>>>> > > > >> > are >> > > >> >> >>>>> > > > >> > >> up >> > > >> >> >>>>> > > > >> > >> >> to >> > > >> >> >>>>> > > > >> > >> >> >>> at >> > > >> >> >>>>> > > > >> > >> >> >>> > any >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > given >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> moment= . >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > it >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = is >> > very >> > > >> possible >> > > >> >> >>>> (given the >> > > >> >> >>>>> > > > existance of >> > > >> >> >>>>> > > > >> > >> headers >> > > >> >> >>>>> > > > >> > >> >> >>> and a >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > correspond= ing >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> plugin >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> > > ecosystem) for >> > > >> >> some >> > > >> >> >>>>> > application to >> > > >> >> >>>>> > > > >> "equip" >> > > >> >> >>>>> > > > >> > >> their >> > > >> >> >>>>> > > > >> > >> >> >>> > >> producers >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > and >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> consumers >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> with >> > > the >> > > >> >> required >> > > >> >> >>>> plugin >> > > >> >> >>>>> > without us >> > > >> >> >>>>> > > > >> > knowing. >> > > >> >> >>>>> > > > >> > >> i >> > > >> >> >>>>> > > > >> > >> >> dont >> > > >> >> >>>>> > > > >> > >> >> >>> > mean >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > to imply >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> thats >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> bad, >> > i >> > > just >> > > >> >> want to >> > > >> >> >>>> make the >> > > >> >> >>>>> > point >> > > >> >> >>>>> > > > that >> > > >> >> >>>>> > > > >> > its >> > > >> >> >>>>> > > > >> > >> not >> > > >> >> >>>>> > > > >> > >> >> as >> > > >> >> >>>>> > > > >> > >> >> >>> > simple >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > keeping it >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> in >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> sync >> > > across a >> > > >> >> >>>> large-enough >> > > >> >> >>>>> > > > organization. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = On >> > > Wed, Nov 9, >> > > >> >> 2016 >> > > >> >> >>>> at 6:17 >> > > >> >> >>>>> > AM, >> > > >> >> >>>>> > > > Magnus >> > > >> >> >>>>> > > > >> > >> Edenhill >> > > >> >> >>>>> > > > >> > >> >> < >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > magnus@edenhill.se> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> > wrote: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > I >> > > think >> > > >> there >> > > >> >> is >> > > >> >> >> a >> > > >> >> >>>> piece >> > > >> >> >>>>> > missing >> > > >> >> >>>>> > > in >> > > >> >> >>>>> > > > >> the >> > > >> >> >>>>> > > > >> > >> >> Strings >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > discussion, >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > where >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> pro-Stringers >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > reason that >> > > >> by >> > > >> >> >>>> providing >> > > >> >> >>>>> > unique >> > > >> >> >>>>> > > > string >> > > >> >> >>>>> > > > >> > >> >> >>> identifiers >> > > >> >> >>>>> > > > >> > >> >> >>> > for >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > each >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > header >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > everything >> > > >> >> will >> > > >> >> >>>> just >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > magically >> > > >> work >> > > >> >> >> for >> > > >> >> >>>> all >> > > >> >> >>>>> > parts of >> > > >> >> >>>>> > > the >> > > >> >> >>>>> > > > >> > stream >> > > >> >> >>>>> > > > >> > >> >> >>> pipeline. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> But >> > > the >> > > >> >> strings >> > > >> >> >>>> dont mean >> > > >> >> >>>>> > > anything >> > > >> >> >>>>> > > > by >> > > >> >> >>>>> > > > >> > >> >> themselves, >> > > >> >> >>>>> > > > >> > >> >> >>> > and >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > while we >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> could >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > probably >> > > >> >> envision >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > some >> > > auto >> > > >> >> plugin >> > > >> >> >>>> loader >> > > >> >> >>>>> > that >> > > >> >> >>>>> > > > >> downloads, >> > > >> >> >>>>> > > > >> > >> >> compiles, >> > > >> >> >>>>> > > > >> > >> >> >>> > links >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > and >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > runs >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> plugins >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > on-demand >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> as >> > > soon as >> > > >> >> >> they're >> > > >> >> >>>> seen by >> > > >> >> >>>>> > a >> > > >> >> >>>>> > > > >> consumer, I >> > > >> >> >>>>> > > > >> > >> dont >> > > >> >> >>>>> > > > >> > >> >> >>> really >> > > >> >> >>>>> > > > >> > >> >> >>> > >> see >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > a >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > use-case >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > for >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > something >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> so >> > > dynamic >> > > >> >> (and >> > > >> >> >>>> fragile) in >> > > >> >> >>>>> > > > practice. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> In >> > > the real >> > > >> >> world >> > > >> >> >>>> an >> > > >> >> >>>>> > application >> > > >> >> >>>>> > > > will >> > > >> >> >>>>> > > > >> be >> > > >> >> >>>>> > > > >> > >> >> >>> configured >> > > >> >> >>>>> > > > >> > >> >> >>> > >> with >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > a set >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > of >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> plugins >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> to >> > > either >> > > >> add >> > > >> >> >>>> (producer) >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> or >> > > read >> > > >> >> >> (consumer) >> > > >> >> >>>> headers. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > This >> > > is an >> > > >> >> >>>> administrative >> > > >> >> >>>>> > task >> > > >> >> >>>>> > > > based >> > > >> >> >>>>> > > > >> on >> > > >> >> >>>>> > > > >> > >> what >> > > >> >> >>>>> > > > >> > >> >> >>> > features a >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > client >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> needs/provides >> > > >> >> >> and >> > > >> >> >>>> results >> > > >> >> >>>>> > in >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > some >> > > sort of >> > > >> >> >>>> configuration >> > > >> >> >>>>> > to >> > > >> >> >>>>> > > > enable >> > > >> >> >>>>> > > > >> and >> > > >> >> >>>>> > > > >> > >> >> >>> configure >> > > >> >> >>>>> > > > >> > >> >> >>> > the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > desired >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> plugins. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > Since this >> > > >> >> needs >> > > >> >> >>>> to be kept >> > > >> >> >>>>> > > > somewhat >> > > >> >> >>>>> > > > >> in >> > > >> >> >>>>> > > > >> > >> sync >> > > >> >> >>>>> > > > >> > >> >> >>> across >> > > >> >> >>>>> > > > >> > >> >> >>> > an >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > organisati= on >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> > (there >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> is >> > > no point >> > > >> in >> > > >> >> >>>> having >> > > >> >> >>>>> > producers >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> add >> > > headers >> > > >> no >> > > >> >> >>>> consumers >> > > >> >> >>>>> > will >> > > >> >> >>>>> > > read, >> > > >> >> >>>>> > > > >> and >> > > >> >> >>>>> > > > >> > >> vice >> > > >> >> >>>>> > > > >> > >> >> >>> versa), >> > > >> >> >>>>> > > > >> > >> >> >>> > >> the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > added >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> > > complexity >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> of >> > > assigning >> > > >> >> an >> > > >> >> >> id >> > > >> >> >>>>> > namespace >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> for >> > > each >> > > >> >> plugin >> > > >> >> >> as >> > > >> >> >>>> it is >> > > >> >> >>>>> > being >> > > >> >> >>>>> > > > >> > configured >> > > >> >> >>>>> > > > >> > >> >> should >> > > >> >> >>>>> > > > >> > >> >> >>> be >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > tolerable. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > /Magnus >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > 2016-11-09 >> > > >> >> 13:06 >> > > >> >> >>>> GMT+01:00 >> > > >> >> >>>>> > > Michael >> > > >> >> >>>>> > > > >> > Pearce < >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > Michael.Pearce@ig.com>: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > Just >> > > >> >> >>>> following/catching >> > > >> >> >>>>> > up on >> > > >> >> >>>>> > > > what >> > > >> >> >>>>> > > > >> > seems >> > > >> >> >>>>> > > > >> > >> to >> > > >> >> >>>>> > > > >> > >> >> be >> > > >> >> >>>>> > > > >> > >> >> >>> an >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > active >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > night :) >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > @Radai >> > > >> >> sorry if >> > > >> >> >>>> it may >> > > >> >> >>>>> > seem >> > > >> >> >>>>> > > > obvious >> > > >> >> >>>>> > > > >> > but >> > > >> >> >>>>> > > > >> > >> what >> > > >> >> >>>>> > > > >> > >> >> >>> does >> > > >> >> >>>>> > > > >> > >> >> >>> > MD >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > stand >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > for? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > My >> > > take on >> > > >> >> >>>> String vs Int: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > I >> > > will >> > > >> state >> > > >> >> >>>> first I am >> > > >> >> >>>>> > pro Int >> > > >> >> >>>>> > > > (16 >> > > >> >> >>>>> > > > >> or >> > > >> >> >>>>> > > > >> > >> 32). >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > I >> > > do >> > > >> though >> > > >> >> >>>> playing >> > > >> >> >>>>> > devils >> > > >> >> >>>>> > > > advocate >> > > >> >> >>>>> > > > >> > see a >> > > >> >> >>>>> > > > >> > >> >> big >> > > >> >> >>>>> > > > >> > >> >> >>> plus >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > with the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> argument >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > of >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > String >> > > >> keys, >> > > >> >> >>>> this is >> > > >> >> >>>>> > around >> > > >> >> >>>>> > > > >> > integrating >> > > >> >> >>>>> > > > >> > >> >> into an >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > existing >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> > eco-system. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > As >> > > many >> > > >> >> other >> > > >> >> >>>> systems use >> > > >> >> >>>>> > > String >> > > >> >> >>>>> > > > >> based >> > > >> >> >>>>> > > > >> > >> >> headers >> > > >> >> >>>>> > > > >> > >> >> >>> > >> (Flume, >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > JMS) >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > it >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > make= s >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = it >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > much >> > > >> easier >> > > >> >> for >> > > >> >> >>>> these to >> > > >> >> >>>>> > be >> > > >> >> >>>>> > > > >> > >> >> >>> > incorporated/integrat= ed >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > into. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > How with >> > > >> Int >> > > >> >> >>>> based >> > > >> >> >>>>> > headers >> > > >> >> >>>>> > > could >> > > >> >> >>>>> > > > we >> > > >> >> >>>>> > > > >> > >> provide >> > > >> >> >>>>> > > > >> > >> >> a >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > way/guidence to >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> make >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > th= is >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > >> integration >> > > >> >> >>>> simple / >> > > >> >> >>>>> > easy with >> > > >> >> >>>>> > > > >> > transition >> > > >> >> >>>>> > > > >> > >> >> flows >> > > >> >> >>>>> > > > >> > >> >> >>> > over >> > > >> >> >>>>> > > > >> > >> >> >>> > >> to >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > kafka? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > * >> > > tough >> > > >> luck >> > > >> >> >>>> buddy >> > > >> >> >>>>> > you're on >> > > >> >> >>>>> > > your >> > > >> >> >>>>> > > > >> own >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > * >> > > simply >> > > >> >> hash >> > > >> >> >>>> the string >> > > >> >> >>>>> > into >> > > >> >> >>>>> > > int >> > > >> >> >>>>> > > > >> code >> > > >> >> >>>>> > > > >> > >> and >> > > >> >> >>>>> > > > >> > >> >> hope >> > > >> >> >>>>> > > > >> > >> >> >>> > for >> > > >> >> >>>>> > > > >> > >> >> >>> > >> no >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > collisions >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > (h= ow >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = to >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > convert >> > > >> back >> > > >> >> >>>> though?) >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > * >> > > http2 >> > > >> >> style >> > > >> >> >> as >> > > >> >> >>>>> > mentioned by >> > > >> >> >>>>> > > > nacho. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > cheers, >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > Mike >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > >> >> >>>>> > ______________________________ >> > > >> >> >>>>> > > > >> > __________ >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > From: >> > > >> radai >> > > >> >> < >> > > >> >> >>>>> > > > >> > radai.rosenblatt@gmail.com> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > Sent: >> > > >> >> >> Wednesday, >> > > >> >> >>>>> > November 9, >> > > >> >> >>>>> > > 2016 >> > > >> >> >>>>> > > > >> > 8:12 AM >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > To: >> > > >> >> >>>> dev@kafka.apache.org >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > Subject: >> > > >> Re: >> > > >> >> >>>> [DISCUSS] >> > > >> >> >>>>> > KIP-82 - >> > > >> >> >>>>> > > > Add >> > > >> >> >>>>> > > > >> > >> Record >> > > >> >> >>>>> > > > >> > >> >> >>> Headers >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > thinking >> > > >> >> about >> > > >> >> >>>> it some >> > > >> >> >>>>> > more, >> > > >> >> >>>>> > > the >> > > >> >> >>>>> > > > >> best >> > > >> >> >>>>> > > > >> > >> way to >> > > >> >> >>>>> > > > >> > >> >> >>> > transmit >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > header >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > remapping >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > data to >> > > >> >> >>>> consumers would >> > > >> >> >>>>> > be to >> > > >> >> >>>>> > > > put it >> > > >> >> >>>>> > > > >> > in >> > > >> >> >>>>> > > > >> > >> the >> > > >> >> >>>>> > > > >> > >> >> MD >> > > >> >> >>>>> > > > >> > >> >> >>> > >> response >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > payload, >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> so >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> maybe >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > it >> > > should >> > > >> be >> > > >> >> >>>> discussed >> > > >> >> >>>>> > now. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > On >> > > Wed, >> > > >> Nov >> > > >> >> 9, >> > > >> >> >>>> 2016 at >> > > >> >> >>>>> > 12:09 >> > > >> >> >>>>> > > AM, >> > > >> >> >>>>> > > > >> > radai < >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> > > >> >> >> radai.rosenblatt@gmail.com >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > wrote: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> > > im not >> > > >> >> >> opposed >> > > >> >> >>>> to the >> > > >> >> >>>>> > idea of >> > > >> >> >>>>> > > > >> > namespace >> > > >> >> >>>>> > > > >> > >> >> >>> mapping. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> all >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > im >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > saying >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> is >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> that >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > its >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> > > not part >> > > >> >> of >> > > >> >> >>>> the "mvp" >> > > >> >> >>>>> > and, >> > > >> >> >>>>> > > > since >> > > >> >> >>>>> > > > >> it >> > > >> >> >>>>> > > > >> > >> >> requires >> > > >> >> >>>>> > > > >> > >> >> >>> no >> > > >> >> >>>>> > > > >> > >> >> >>> > >> wire >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > format >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > chan= ge, >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = can >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> > > always >> > > >> be >> > > >> >> >>>> added later. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> > > also, >> > > >> its >> > > >> >> not >> > > >> >> >>>> as >> > > >> >> >>>>> > simple as >> > > >> >> >>>>> > > just >> > > >> >> >>>>> > > > >> > >> >> configuring >> > > >> >> >>>>> > > > >> > >> >> >>> MM >> > > >> >> >>>>> > > > >> > >> >> >>> > to >> > > >> >> >>>>> > > > >> > >> >> >>> > >> do >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> > transform: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > lets >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> > > say i've >> > > >> >> >>>> implemented >> > > >> >> >>>>> > large >> > > >> >> >>>>> > > > message >> > > >> >> >>>>> > > > >> > >> >> support as >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > {666,1} and >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > on >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> some >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > mirror >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> > > target >> > > >> >> >> cluster >> > > >> >> >>>> its been >> > > >> >> >>>>> > > > remapped >> > > >> >> >>>>> > > > >> to >> > > >> >> >>>>> > > > >> > >> >> {999,1}. >> > > >> >> >>>>> > > > >> > >> >> >>> the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > consumer >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> plugin >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > co= de >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > would >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> > > also >> > > >> need >> > > >> >> to >> > > >> >> >>>> be told >> > > >> >> >>>>> > to look >> > > >> >> >>>>> > > > for >> > > >> >> >>>>> > > > >> the >> > > >> >> >>>>> > > > >> > >> large >> > > >> >> >>>>> > > > >> > >> >> >>> > message >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > "part X >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > of >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> Y" >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> > header >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> > > under >> > > >> >> >> {999,1}. >> > > >> >> >>>> doable, >> > > >> >> >>>>> > but >> > > >> >> >>>>> > > > tricky. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> > > On Tue, >> > > >> >> Nov >> > > >> >> >> 8, >> > > >> >> >>>> 2016 at >> > > >> >> >>>>> > 10:29 >> > > >> >> >>>>> > > > PM, >> > > >> >> >>>>> > > > >> > Gwen >> > > >> >> >>>>> > > > >> > >> >> >>> Shapira < >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> >> > > gwen@confluent.io >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > wrote: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > While >> > > >> you >> > > >> >> >> can >> > > >> >> >>>> do >> > > >> >> >>>>> > whatever >> > > >> >> >>>>> > > you >> > > >> >> >>>>> > > > >> want >> > > >> >> >>>>> > > > >> > >> with a >> > > >> >> >>>>> > > > >> > >> >> >>> > >> namespace >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > and >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > your >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > code= , >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > what >> > > >> I'd >> > > >> >> >>>> expect is >> > > >> >> >>>>> > for each >> > > >> >> >>>>> > > > app >> > > >> >> >>>>> > > > >> to >> > > >> >> >>>>> > > > >> > >> >> >>> namespaces >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> configurable... >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > So if I >> > > >> >> >>>> accidentally >> > > >> >> >>>>> > used >> > > >> >> >>>>> > > 666 >> > > >> >> >>>>> > > > for >> > > >> >> >>>>> > > > >> > my >> > > >> >> >>>>> > > > >> > >> HR >> > > >> >> >>>>> > > > >> > >> >> >>> > >> department, >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > and >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > still >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > want >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = to >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > run >> > > >> >> >> RadaiApp, >> > > >> >> >>>> I can >> > > >> >> >>>>> > config >> > > >> >> >>>>> > > > >> > >> "namespace=3D42" >> > > >> >> >>>>> > > > >> > >> >> >>> for >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > RadaiApp and >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> > > everything >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > will >> > > >> look >> > > >> >> >>>> normal. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > This >> > > >> >> means >> > > >> >> >>>> you only >> > > >> >> >>>>> > need to >> > > >> >> >>>>> > > > sync >> > > >> >> >>>>> > > > >> > usage >> > > >> >> >>>>> > > > >> > >> >> >>> inside >> > > >> >> >>>>> > > > >> > >> >> >>> > your >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > own >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> > > organization. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > Still >> > > >> >> hard, >> > > >> >> >>>> but >> > > >> >> >>>>> > somewhat >> > > >> >> >>>>> > > > easier >> > > >> >> >>>>> > > > >> > than >> > > >> >> >>>>> > > > >> > >> >> syncing >> > > >> >> >>>>> > > > >> > >> >> >>> > with >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > entire >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> world. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > On Tue, >> > > >> >> Nov >> > > >> >> >>>> 8, 2016 >> > > >> >> >>>>> > at 10:07 >> > > >> >> >>>>> > > > PM, >> > > >> >> >>>>> > > > >> > >> radai < >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> > > >> >> >>>> radai.rosenblatt@gmail.com> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > wrote: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > > and >> > > >> we >> > > >> >> can >> > > >> >> >>>> start >> > > >> >> >>>>> > with >> > > >> >> >>>>> > > > >> {namespace, >> > > >> >> >>>>> > > > >> > >> id} >> > > >> >> >>>>> > > > >> > >> >> and >> > > >> >> >>>>> > > > >> > >> >> >>> no >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > re-mapping >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > supp= ort >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = and >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > always >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > > add >> > > >> it >> > > >> >> >>>> later on >> > > >> >> >>>>> > if/when >> > > >> >> >>>>> > > > >> > collisions >> > > >> >> >>>>> > > > >> > >> >> >>> actually >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > happen (i >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > dont >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > thin= k >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > they'd >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > be >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > > a >> > > >> >> >> problem). >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > > every >> > > >> >> >>>> interested >> > > >> >> >>>>> > party (so >> > > >> >> >>>>> > > > orgs >> > > >> >> >>>>> > > > >> > or >> > > >> >> >>>>> > > > >> > >> >> >>> > individuals) >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > could >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > then >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> > > register >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > a >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > > >> > > >> prefix >> > > >> >> (0 >> > > >> >> >> =3D >> > > >> >> >>>>> > reserved, 1 =3D >> > > >> >> >>>>> > > > >> > confluent >> > > >> >> >>>>> > > > >> > >> ... >> > > >> >> >>>>> > > > >> > >> >> >>> 666 >> > > >> >> >>>>> > > > >> > >> >> >>> > =3D me >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > :-) ) >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > and >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> do >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > whatever >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > with >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > > the >> > > >> >> 2nd ID >> > > >> >> >>>> - so once >> > > >> >> >>>>> > > > linkedin >> > > >> >> >>>>> > > > >> > >> >> registers, >> > > >> >> >>>>> > > > >> > >> >> >>> say >> > > >> >> >>>>> > > > >> > >> >> >>> > 3, >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > then >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> linked= in >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > de= vs >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> are >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > free >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > > to >> > > >> use >> > > >> >> {3, >> > > >> >> >>>> *} with a >> > > >> >> >>>>> > > > reasonable >> > > >> >> >>>>> > > > >> > >> >> >>> expectation >> > > >> >> >>>>> > > > >> > >> >> >>> > to >> > > >> >> >>>>> > > > >> > >> >> >>> > >> to >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > collide >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> with >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > anything >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > > else. >> > > >> >> >>>> further >> > > >> >> >>>>> > partitioning >> > > >> >> >>>>> > > > of >> > > >> >> >>>>> > > > >> > that * >> > > >> >> >>>>> > > > >> > >> >> >>> becomes >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > linkedin's >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> problem, >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = but >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > > >> > > >> >> "upstream >> > > >> >> >>>>> > registration" >> > > >> >> >>>>> > > of a >> > > >> >> >>>>> > > > >> > >> namespace >> > > >> >> >>>>> > > > >> > >> >> >>> only >> > > >> >> >>>>> > > > >> > >> >> >>> > has >> > > >> >> >>>>> > > > >> > >> >> >>> > >> to >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > happen >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > once= . >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > > On >> > > >> Tue, >> > > >> >> >> Nov >> > > >> >> >>>> 8, 2016 >> > > >> >> >>>>> > at >> > > >> >> >>>>> > > 9:03 >> > > >> >> >>>>> > > > PM, >> > > >> >> >>>>> > > > >> > >> James >> > > >> >> >>>>> > > > >> > >> >> >>> Cheng < >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> > > >> >> wushujames@gmail.com >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > wrote: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> > On >> > > >> >> Nov >> > > >> >> >>>> 8, 2016, >> > > >> >> >>>>> > at 5:54 >> > > >> >> >>>>> > > > PM, >> > > >> >> >>>>> > > > >> > Gwen >> > > >> >> >>>>> > > > >> > >> >> >>> Shapira < >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > >> > > gwen@confluent.io >> > > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > wrote: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> > >> > > >> >> Thank >> > > >> >> >>>> you so >> > > >> >> >>>>> > much for >> > > >> >> >>>>> > > > this >> > > >> >> >>>>> > > > >> > clear >> > > >> >> >>>>> > > > >> > >> and >> > > >> >> >>>>> > > > >> > >> >> >>> fair >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > summary of >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > arguments. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> > >> > > >> I'm >> > > >> >> in >> > > >> >> >>>> favor of >> > > >> >> >>>>> > ints. >> > > >> >> >>>>> > > > Not a >> > > >> >> >>>>> > > > >> > >> >> >>> deal-breaker, >> > > >> >> >>>>> > > > >> > >> >> >>> > but >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > in >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > favor. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> > >> > > >> Even >> > > >> >> >>>> more in >> > > >> >> >>>>> > favor of >> > > >> >> >>>>> > > > >> Magnus's >> > > >> >> >>>>> > > > >> > >> >> >>> > decentralized >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > suggestion >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > with >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > Roger's >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> > >> > > >> >> tweak: >> > > >> >> >>>> add a >> > > >> >> >>>>> > namespace >> > > >> >> >>>>> > > > for >> > > >> >> >>>>> > > > >> > >> headers. >> > > >> >> >>>>> > > > >> > >> >> >>> This >> > > >> >> >>>>> > > > >> > >> >> >>> > will >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > allow >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > each >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > app >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > to >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > just >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> > >> > > >> use >> > > >> >> >>>> whatever IDs >> > > >> >> >>>>> > it >> > > >> >> >>>>> > > wants >> > > >> >> >>>>> > > > >> > >> >> internally, >> > > >> >> >>>>> > > > >> > >> >> >>> and >> > > >> >> >>>>> > > > >> > >> >> >>> > >> then >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > let >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > admi= n >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> deploying >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> > >> > > >> the >> > > >> >> app >> > > >> >> >>>> figure >> > > >> >> >>>>> > out an >> > > >> >> >>>>> > > > >> > available >> > > >> >> >>>>> > > > >> > >> >> >>> namespace >> > > >> >> >>>>> > > > >> > >> >> >>> > ID >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > for the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > app >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> to >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> live >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > in. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> > So >> > > >> >> >>>>> > > > >> > io.confluent.schema-registry >> > > >> >> >>>>> > > > >> > >> can >> > > >> >> >>>>> > > > >> > >> >> be >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > namespace >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > 0x01 on >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> my >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> deployment >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> > >> > > >> and >> > > >> >> >> 0x57 >> > > >> >> >>>> on >> > > >> >> >>>>> > yours, and >> > > >> >> >>>>> > > > the >> > > >> >> >>>>> > > > >> > poor >> > > >> >> >>>>> > > > >> > >> guys >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > developing the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > app >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > don'= t >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > need >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > to >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> > >> > > >> >> worry >> > > >> >> >>>> about that. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> > > >> Gwen, >> > > >> >> if >> > > >> >> >> I >> > > >> >> >>>>> > understand >> > > >> >> >>>>> > > your >> > > >> >> >>>>> > > > >> > example >> > > >> >> >>>>> > > > >> > >> >> >>> right, an >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > applicatio= n >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > >> > deployer >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > might >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> > > >> >> decide to >> > > >> >> >>>> use 0x01 >> > > >> >> >>>>> > in one >> > > >> >> >>>>> > > > >> > >> deployment, >> > > >> >> >>>>> > > > >> > >> >> and >> > > >> >> >>>>> > > > >> > >> >> >>> > that >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > means >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > that >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > once >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > message >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> is >> > > >> >> >> written >> > > >> >> >>>> into the >> > > >> >> >>>>> > > > broker, it >> > > >> >> >>>>> > > > >> > >> will be >> > > >> >> >>>>> > > > >> > >> >> >>> > saved on >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > broker >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > with >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > that >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> > > >> >> specific >> > > >> >> >>>> namespace >> > > >> >> >>>>> > > (0x01). >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> If >> > > >> you >> > > >> >> >>>> were to >> > > >> >> >>>>> > mirror >> > > >> >> >>>>> > > that >> > > >> >> >>>>> > > > >> > message >> > > >> >> >>>>> > > > >> > >> >> into >> > > >> >> >>>>> > > > >> > >> >> >>> > another >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > cluster, >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > 0x= 01 >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > would >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> > > >> >> accompany >> > > >> >> >>>> the >> > > >> >> >>>>> > message, >> > > >> >> >>>>> > > > right? >> > > >> >> >>>>> > > > >> > What >> > > >> >> >>>>> > > > >> > >> if >> > > >> >> >>>>> > > > >> > >> >> the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > deployers of >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > same >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = app >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > in >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> > > >> other >> > > >> >> >>>> cluster uses >> > > >> >> >>>>> > 0x57? >> > > >> >> >>>>> > > > They >> > > >> >> >>>>> > > > >> > won't >> > > >> >> >>>>> > > > >> > >> >> >>> > understand >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > each >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > other? >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> I'm >> > > >> >> not >> > > >> >> >>>> sure >> > > >> >> >>>>> > that's an >> > > >> >> >>>>> > > > >> avoidable >> > > >> >> >>>>> > > > >> > >> >> >>> problem. I >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > think it >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > simply >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > me= ans >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > that >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > in >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> > > >> order >> > > >> >> to >> > > >> >> >>>> share >> > > >> >> >>>>> > data, you >> > > >> >> >>>>> > > > have >> > > >> >> >>>>> > > > >> to >> > > >> >> >>>>> > > > >> > >> also >> > > >> >> >>>>> > > > >> > >> >> >>> have a >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > shared >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > (agreed >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > up= on) >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> > > >> >> >>>> understanding of >> > > >> >> >>>>> > what the >> > > >> >> >>>>> > > > >> > >> namespaces >> > > >> >> >>>>> > > > >> > >> >> >>> mean. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> Which >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > I >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > think >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > make= s >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > sense, >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> > > >> >> because >> > > >> >> >> the >> > > >> >> >>>>> > alternate >> > > >> >> >>>>> > > > (sharing >> > > >> >> >>>>> > > > >> > >> >> *nothing* >> > > >> >> >>>>> > > > >> > >> >> >>> at >> > > >> >> >>>>> > > > >> > >> >> >>> > >> all) >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > would >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > mean >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > th= at >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > > there >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> > > >> would >> > > >> >> be >> > > >> >> >>>> no way to >> > > >> >> >>>>> > > > understand >> > > >> >> >>>>> > > > >> > each >> > > >> >> >>>>> > > > >> > >> >> other. >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> > > >> -James >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> > >> > > >> Gwen >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> > On >> > > >> >> Tue, >> > > >> >> >>>> Nov 8, >> > > >> >> >>>>> > 2016 at >> > > >> >> >>>>> > > > 4:23 >> > > >> >> >>>>> > > > >> > PM, >> > > >> >> >>>>> > > > >> > >> >> radai < >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > >> > > >> >> >>>> radai.rosenblatt@gmail.com >> > > >> >> >>>>> > > >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> > > >> wrote: >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> >> > > >> +1 >> > > >> >> for >> > > >> >> >>>> sean's >> > > >> >> >>>>> > > document. >> > > >> >> >>>>> > > > it >> > > >> >> >>>>> > > > >> > >> covers >> > > >> >> >>>>> > > > >> > >> >> >>> pretty >> > > >> >> >>>>> > > > >> > >> >> >>> > >> much >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > all >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > the >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > >> > > trade-offs >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > and >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> >> > > >> >> >> provides >> > > >> >> >>>>> > concrete >> > > >> >> >>>>> > > > figures >> > > >> >> >>>>> > > > >> to >> > > >> >> >>>>> > > > >> > >> argue >> > > >> >> >>>>> > > > >> > >> >> >>> about >> > > >> >> >>>>> > > > >> > >> >> >>> > :-) >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> >> > > >> >> >>>> (nit-picking - >> > > >> >> >>>>> > used >> > > >> >> >>>>> > > the >> > > >> >> >>>>> > > > >> same >> > > >> >> >>>>> > > > >> > >> xkcd >> > > >> >> >>>>> > > > >> > >> >> >>> twice, >> > > >> >> >>>>> > > > >> > >> >> >>> > >> also >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > trove >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > has >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > been >> > > >> >> >>>>> > > > >> > >> >> >>> > >> > > >> > > > = > > >> > >> >> > > >> >> superceded >> > > >> >> >>>>> > > > >> > >> > >> > > >> >> >>>>> > > > >> > >> > > >> >> >>>>> > > > >> > >> > > >> >> >>>>> > > > >> > >> > > >> >> >>>>> > > > >> > -- >> > > >> >> >>>>> > > > >> > Gwen Shapira >> > > >> >> >>>>> > > > >> > Product Manager | Confluent >> > > >> >> >>>>> > > > >> > 650.450.2760 <(650)%20450-2760> | >> > > @gwenshap >> > > >> >> >>>>> > > > >> > Follow us: Twitter | blog >> > > >> >> >>>>> > > > >> > >> > > >> >> >>>>> > > > >> >> > > >> >> >>>>> > > > >> >> > > >> >> >>>>> > > > >> >> > > >> >> >>>>> > > > >> -- >> > > >> >> >>>>> > > > >> *Todd Palino* >> > > >> >> >>>>> > > > >> Staff Site Reliability Engineer >> > > >> >> >>>>> > > > >> Data Infrastructure Streaming >> > > >> >> >>>>> > > > >> >> > > >> >> >>>>> > > > >> >> > > >> >> >>>>> > > > >> >> > > >> >> >>>>> > > > >> linkedin.com/in/toddpalino >> > > >> >> >>>>> > > > >> >> > > >> >> >>>>> > > > >> > > >> >> >>>>> > > > >> > > >> >> >>>>> > > > >> > > >> >> >>>>> > > > -- >> > > >> >> >>>>> > > > Gwen Shapira >> > > >> >> >>>>> > > > Product Manager | Confluent >> > > >> >> >>>>> > > > 650.450.2760 <(650)%20450-2760> | >> @gwenshap >> > > >> >> >>>>> > > > Follow us: Twitter | blog >> > > >> >> >>>>> > > > >> > > >> >> >>>>> > > >> > > >> >> >>>>> > >> > > >> >> >>>>> > >> > > >> >> >>>>> > The information contained in this email is >> strictly >> > > >> >> confidential >> > > >> >> >>>> and for >> > > >> >> >>>>> > the use of the addressee only, unless otherwis= e >> > > indicated. >> > > >> If >> > > >> >> you >> > > >> >> >>>> are not >> > > >> >> >>>>> > the intended recipient, please do not read, co= py, >> > > use or >> > > >> >> disclose >> > > >> >> >>>> to others >> > > >> >> >>>>> > this message or any attachment. Please also >> notify >> > > the >> > > >> sender >> > > >> >> by >> > > >> >> >>>> replying >> > > >> >> >>>>> > to this email or by telephone (+44(020 7896 00= 11) >> > > and then >> > > >> >> delete >> > > >> >> >>>> the email >> > > >> >> >>>>> > and any copies of it. Opinions, conclusion (et= c) >> > > that do not >> > > >> >> >>>> relate to the >> > > >> >> >>>>> > official business of this company shall be >> > > understood as >> > > >> >> neither >> > > >> >> >>>> given nor >> > > >> >> >>>>> > endorsed by it. IG is a trading name of IG >> Markets >> > > Limited >> > > >> (a >> > > >> >> >>>> company >> > > >> >> >>>>> > registered in England and Wales, company numbe= r >> > > 04008957) >> > > >> and >> > > >> >> IG >> > > >> >> >>>> Index >> > > >> >> >>>>> > Limited (a company registered in England and >> Wales, >> > > company >> > > >> >> >> number >> > > >> >> >>>>> > 01190902). Registered address at Cannon Bridge >> > > House, 25 >> > > >> >> Dowgate >> > > >> >> >>>> Hill, >> > > >> >> >>>>> > London EC4R 2YA. Both IG Markets Limited >> (register >> > > number >> > > >> >> 195355) >> > > >> >> >>>> and IG >> > > >> >> >>>>> > Index Limited (register number 114059) are >> > > authorised and >> > > >> >> >>>> regulated by the >> > > >> >> >>>>> > Financial Conduct Authority. >> > > >> >> >>>>> > >> > > >> >> >>>>> >> > > >> >> >>>>> >> > > >> >> >>>>> The information contained in this email is strictly >> > > confidential >> > > >> and >> > > >> >> >> for >> > > >> >> >>>> the use of the addressee only, unless otherwise >> indicated. >> > > If you >> > > >> are >> > > >> >> >> not >> > > >> >> >>>> the intended recipient, please do not read, copy, use= or >> > > disclose >> > > >> to >> > > >> >> >> others >> > > >> >> >>>> this message or any attachment. Please also notify th= e >> > > sender by >> > > >> >> >> replying >> > > >> >> >>>> to this email or by telephone (+44(020 7896 0011) and >> then >> > > delete >> > > >> the >> > > >> >> >> email >> > > >> >> >>>> and any copies of it. Opinions, conclusion (etc) that= do >> > > not relate >> > > >> >> to >> > > >> >> >> the >> > > >> >> >>>> official business of this company shall be understood= as >> > > neither >> > > >> >> given >> > > >> >> >> nor >> > > >> >> >>>> endorsed by it. IG is a trading name of IG Markets >> Limited >> > > (a >> > > >> company >> > > >> >> >>>> registered in England and Wales, company number >> 04008957) >> > > and IG >> > > >> >> Index >> > > >> >> >>>> Limited (a company registered in England and Wales, >> > company >> > > number >> > > >> >> >>>> 01190902). Registered address at Cannon Bridge House,= 25 >> > > Dowgate >> > > >> >> Hill, >> > > >> >> >>>> London EC4R 2YA. Both IG Markets Limited (register >> number >> > > 195355) >> > > >> >> and IG >> > > >> >> >>>> Index Limited (register number 114059) are authorised >> and >> > > regulated >> > > >> >> by >> > > >> >> >> the >> > > >> >> >>>> Financial Conduct Authority. >> > > >> >> >>>>> >> > > >> >> >>>> >> > > >> >> >>>> >> > > >> >> >>> >> > > >> >> >> >> > > >> >> >> >> > > >> >> > >> > > >> >> >> > > >> >> >> > > >> > >> > > >> The information contained in this email is strictly confident= ial >> > > and for >> > > >> the use of the addressee only, unless otherwise indicated. If >> you >> > > are not >> > > >> the intended recipient, please do not read, copy, use or >> disclose >> > > to others >> > > >> this message or any attachment. Please also notify the sender= by >> > > replying >> > > >> to this email or by telephone (+44(020 7896 0011) and then >> delete >> > > the email >> > > >> and any copies of it. Opinions, conclusion (etc) that do not >> > relate >> > > to the >> > > >> official business of this company shall be understood as neit= her >> > > given nor >> > > >> endorsed by it. IG is a trading name of IG Markets Limited (a >> > > company >> > > >> registered in England and Wales, company number 04008957) and= IG >> > > Index >> > > >> Limited (a company registered in England and Wales, company >> number >> > > >> 01190902). Registered address at Cannon Bridge House, 25 Dowg= ate >> > > Hill, >> > > >> London EC4R 2YA. Both IG Markets Limited (register number >> 195355) >> > > and IG >> > > >> Index Limited (register number 114059) are authorised and >> > regulated >> > > by the >> > > >> Financial Conduct Authority. >> > > >> >> > > >> > > >> > > >> > > -- >> > > Gwen Shapira >> > > Product Manager | Confluent >> > > 650.450.2760 | @gwenshap >> > > Follow us: Twitter | blog >> > > >> > > >> > > The information contained in this email is strictly confidential and >> for >> > > the use of the addressee only, unless otherwise indicated. If you ar= e >> not >> > > the intended recipient, please do not read, copy, use or disclose to >> > others >> > > this message or any attachment. Please also notify the sender by >> replying >> > > to this email or by telephone (+44(020 7896 0011) and then delete th= e >> > email >> > > and any copies of it. Opinions, conclusion (etc) that do not relate = to >> > the >> > > official business of this company shall be understood as neither giv= en >> > nor >> > > endorsed by it. IG is a trading name of IG Markets Limited (a compan= y >> > > registered in England and Wales, company number 04008957) and IG Ind= ex >> > > Limited (a company registered in England and Wales, company number >> > > 01190902). Registered address at Cannon Bridge House, 25 Dowgate Hil= l, >> > > London EC4R 2YA. Both IG Markets Limited (register number 195355) an= d >> IG >> > > Index Limited (register number 114059) are authorised and regulated = by >> > the >> > > Financial Conduct Authority. >> > > >> > >>