From dev-return-12166-archive-asf-public=cust-asf.ponee.io@beam.apache.org Thu Sep 13 07:13:43 2018 Return-Path: X-Original-To: archive-asf-public@cust-asf.ponee.io Delivered-To: archive-asf-public@cust-asf.ponee.io Received: from mail.apache.org (hermes.apache.org [140.211.11.3]) by mx-eu-01.ponee.io (Postfix) with SMTP id 2D93F180649 for ; Thu, 13 Sep 2018 07:13:42 +0200 (CEST) Received: (qmail 92151 invoked by uid 500); 13 Sep 2018 05:13:41 -0000 Mailing-List: contact dev-help@beam.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Reply-To: dev@beam.apache.org Delivered-To: mailing list dev@beam.apache.org Received: (qmail 92139 invoked by uid 99); 13 Sep 2018 05:13:40 -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; Thu, 13 Sep 2018 05:13:40 +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 E811C1A127D for ; Thu, 13 Sep 2018 05:13:39 +0000 (UTC) X-Virus-Scanned: Debian amavisd-new at spamd2-us-west.apache.org X-Spam-Flag: NO X-Spam-Score: 1.869 X-Spam-Level: * X-Spam-Status: No, score=1.869 tagged_above=-999 required=6.31 tests=[DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, HTML_MESSAGE=2, RCVD_IN_DNSWL_NONE=-0.0001, RCVD_IN_MSPIKE_H3=-0.01, RCVD_IN_MSPIKE_WL=-0.01, SPF_PASS=-0.001, T_DKIMWL_WL_MED=-0.01] autolearn=disabled Authentication-Results: spamd2-us-west.apache.org (amavisd-new); dkim=pass (2048-bit key) header.d=gmail.com Received: from mx1-lw-us.apache.org ([10.40.0.8]) by localhost (spamd2-us-west.apache.org [10.40.0.9]) (amavisd-new, port 10024) with ESMTP id 21m6myhlkBqf for ; Thu, 13 Sep 2018 05:13:37 +0000 (UTC) Received: from mail-yw1-f44.google.com (mail-yw1-f44.google.com [209.85.161.44]) by mx1-lw-us.apache.org (ASF Mail Server at mx1-lw-us.apache.org) with ESMTPS id 714575F3B3 for ; Thu, 13 Sep 2018 05:13:37 +0000 (UTC) Received: by mail-yw1-f44.google.com with SMTP id l9-v6so677421ywc.11 for ; Wed, 12 Sep 2018 22:13:37 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=mime-version:references:in-reply-to:from:date:message-id:subject:to; bh=RPGjnNiX/JvhzUryqa9UG6sTKjUfr59X3z7z/1j8+UY=; b=jQC15S8TPjH4GKSqX7/i2iEMmxSA3BXgTkV90KZCoPmqWiwGW2X6kVqqANgO22yJcB Scs3v8GEN1R69LyOXgKzshv/0OqNSj8icBUTmnrj75SZS40JlLMv9fraouJ/2t9uqf6Z VMR8L+j9N+X1UcUNwRBhPFhXiE/9BWsctCzPMmJE7jT/ANb6/NI5aDG0uU+fpi2kIqJV fT2vtbr6bg8mcxsYM/DAYLIhWuSj1z2ODqXwf3eEb5XzVSr7gMgJlnldXBvmUEgcekfJ /kZRZ95xJ00DWZG3MEFjs810046oVBOuv04DA+8fP2R8A3dUOaseLX13ciqKwJvvMDst Dniw== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to; bh=RPGjnNiX/JvhzUryqa9UG6sTKjUfr59X3z7z/1j8+UY=; b=IeFKI6MPikj8gMgDk2q4Vy7UZJjOZPAE1RLRz7+Gs+rqk36e/ggOWl1zErEvdnJc+N CF8bKc+TnZ4z7kSHrKJWnW7KaieEMvXoImMfB0K01BTTBWAf+s/zD/A4ZWJVuexNJQhv x3OCOYzo4t7CVl0Nr4HV2erkf+78WaFtSVL5ChfQATtzLWQbNd4/QEhoFgZmnt7x1t+r qrM6MuemkSrgnu9cml4/DfvSGcPhimE/2/KTZ/1Vtej1VyMTMfoSL6Le/xzTB6mN2WA1 AAL0ZTgYOkV+Uz7+pD6ySIBHlCI7kFAtlUMdBsa0gKcl9P7XB6VtI9wvyC/Smd9WpsnC FQSw== X-Gm-Message-State: APzg51BRwbIP0gOYO5lMvJxK5/p5gvk6Eq42LA9DmhQEewc7vf7xMTZt zK6Rny0FeC+kLOHFX7LxmZ6IVDKJ7Um94sx0GDMj6g== X-Google-Smtp-Source: ANB0VdZki3NW6GcZENZ7Fb7ww9os98IzR+Eq5utwHiOY1jXJT0KATiLDfpP3Q0y1FxGOp6PvF06h2iUl+x7ZuEg0RdI= X-Received: by 2002:a81:9e58:: with SMTP id v85-v6mr2679353ywg.343.1536815610691; Wed, 12 Sep 2018 22:13:30 -0700 (PDT) MIME-Version: 1.0 References: <5d5ab126-c296-c7e5-d0d4-f83048c12ed2@apache.org> In-Reply-To: From: Romain Manni-Bucau Date: Thu, 13 Sep 2018 07:13:17 +0200 Message-ID: Subject: Re: [Discuss] Upgrade story for Beam's execution engines To: dev@beam.apache.org Content-Type: multipart/alternative; boundary="0000000000008f87a50575b9c466" --0000000000008f87a50575b9c466 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Hi guys, Isnt the issue "only" that beam has this code instead of engines? Assuming beam runner facing api is stable - which must be the case anyway - and that each engine has its integration (flink-beam instead of beam-runners-flink), then this issue disappears by construction. It also has the advantage to have a better maintenance. Side note: this is what happent which arquillian, originally the community did all adapters impl then each vendor took it back in house to make it better. Any way to work in that direction maybe? Le jeu. 13 sept. 2018 00:49, Thomas Weise a =C3=A9crit : > The main problem here is that users are forced to upgrade infrastructure > to obtain new features in Beam, even when those features actually don't > require such changes. As an example, another update to Flink 1.6.0 was > proposed (without supporting new functionality in Beam) and we already kn= ow > that it breaks compatibility (again). > > I think that upgrading to a Flink X.Y.0 version isn't a good idea to star= t > with. But besides that, if we want to grow adoption, then we need to focu= s > on stability and delivering improvements to Beam without disrupting users= . > > In the specific case, ideally the surface of Flink would be backward > compatible, allowing us to stick to a minimum version and be able to subm= it > pipelines to Flink endpoints of higher versions. Some work in that > direction is underway (like versioning the REST API). FYI, lowest common > version is what most projects that depend on Hadoop 2.x follow. > > Since Beam with Flink 1.5.x client won't talk to Flink 1.6 and there are > code changes required to make it compile, we would need to come up with a > more involved strategy to support multiple Flink versions. Till then, I > would prefer we favor existing users over short lived experiments, which > would mean stick with 1.5.x and not support 1.6.0. > > Thanks, > Thomas > > > On Wed, Sep 12, 2018 at 1:15 PM Lukasz Cwik wrote: > >> As others have already suggested, I also believe LTS releases is the bes= t >> we can do as a community right now until portability allows us to decoup= le >> what a user writes with and how it runs (the SDK and the SDK environment= ) >> from the runner (job service + shared common runner libs + >> Flink/Spark/Dataflow/Apex/Samza/...). >> >> Dataflow would be highly invested in having the appropriate tooling >> within Apache Beam to support multiple SDK versions against a runner. Th= is >> in turn would allow people to use any SDK with any runner and as Robert = had >> mentioned, certain optimizations and features would be disabled dependin= g >> on the capabilities of the runner and the capabilities of the SDK. >> >> >> >> On Wed, Sep 12, 2018 at 6:38 AM Robert Bradshaw >> wrote: >> >>> The target audience is people who want to use the latest Beam but do no= t >>> want to use the latest version of the runner, right? >>> >>> I think this will be somewhat (though not entirely) addressed by Beam >>> LTS releases, where those not wanting to upgrade the runner at least ha= ve a >>> well-supported version of Beam. In the long term, we have the division >>> >>> Runner <-> BeamRunnerSpecificCode <-> CommonBeamRunnerLibs <-> SDK. >>> >>> (which applies to the job submission as well as execution). >>> >>> Insomuch as the BeamRunnerSpecificCode uses the public APIs of the >>> runner, hopefully upgrading the runner for minor versions should be a >>> no-op, and we can target the lowest version of the runner that makes se= nse, >>> allowing the user to link against higher versions at his or her discret= ion. >>> We should provide built targets that allow this. For major versions, it= may >>> make sense to have two distinct BeamRunnerSpecificCode libraries (which= may >>> or may not share some common code). I hope these wrappers are not too >>> thick. >>> >>> There is a tight coupling at the BeamRunnerSpecificCode <-> >>> CommonBeamRunnerLibs layer, but hopefully the bulk of the code lives on= the >>> right hand side and can be updated as needed independent of the runner. >>> There may be code of the form "if the runner supports X, do this fast p= ath, >>> otherwise, do this slow path (or reject the pipeline). >>> >>> I hope the CommonBeamRunnerLibs <-> SDK coupling is fairly loose, to th= e >>> point that one could use SDKs from different versions of Beam (or even >>> developed outside of Beam) with an older/newer runner. We may need to a= dd >>> versioning to the Fn/Runner/Job API itself to support this. Right now o= f >>> course we're still in a pre-1.0, rapid-development phase wrt this API. >>> >>> >>> >>> >>> On Wed, Sep 12, 2018 at 2:10 PM Etienne Chauchot >>> wrote: >>> >>>> Hi Max, >>>> >>>> I totally agree with your points especially the users priorities (stic= k >>>> to the already working version) , and the need to leverage important n= ew >>>> features. It is indeed a difficult balance to find . >>>> >>>> I can talk for a part I know: for the Spark runner, the aim was to >>>> support Dataset native spark API (in place of RDD). For that we needed= to >>>> upgrade to spark 2.x (and we will probably leverage Beam Row as well). >>>> But such an upgrade is a good amount of work which makes it difficult >>>> to commit on a schedule such as "if there is a major new feature on an >>>> execution engine that we want to leverage, then the upgrade in Beam wi= ll be >>>> done within x months". >>>> >>>> Regarding your point on portability : decoupling SDK from runner with >>>> runner harness and SDK harness might make pipeline authors work easy >>>> regarding pipeline maintenance. But, still, if we upgrade runner libs,= then >>>> the users might have their runner harness not work with their engine >>>> version. >>>> If such SDK/runner decoupling is 100% functional, then we could imagin= g >>>> having multiple runner harnesses shipping different versions of the ru= nner >>>> libs to solve this problem. >>>> But we would need to support more than one version of the runner libs. >>>> We chose not to do this on spark runner. >>>> >>>> WDYT ? >>>> >>>> Best >>>> Etienne >>>> >>>> >>>> Le mardi 11 septembre 2018 =C3=A0 15:42 +0200, Maximilian Michels a = =C3=A9crit : >>>> >>>> Hi Beamers, >>>> >>>> >>>> In the light of the discussion about Beam LTS releases, I'd like to ki= ck >>>> >>>> off a thread about how often we upgrade the execution engine of each >>>> >>>> Runner. By upgrade, I mean major/minor versions which typically break >>>> >>>> the binary compatibility of Beam pipelines. >>>> >>>> >>>> For the Flink Runner, we try to track the latest stable version. Some >>>> >>>> users reported that this can be problematic, as it requires them to >>>> >>>> potentially upgrade their Flink cluster with a new version of Beam. >>>> >>>> >>>> From a developer's perspective, it makes sense to migrate as early as >>>> >>>> possible to the newest version of the execution engine, e.g. to levera= ge >>>> >>>> the newest features. From a user's perspective, you don't care about t= he >>>> >>>> latest features if your use case still works with Beam. >>>> >>>> >>>> We have to please both parties. So I'd suggest to upgrade the executio= n >>>> >>>> engine whenever necessary (e.g. critical new features, end of life of >>>> >>>> current version). On the other hand, the upcoming Beam LTS releases wi= ll >>>> >>>> contain a longer-supported version. >>>> >>>> >>>> Maybe we don't need to discuss much about this but I wanted to hear wh= at >>>> >>>> the community has to say about it. Particularly, I'd be interested in >>>> >>>> how the other Runner authors intend to do it. >>>> >>>> >>>> As far as I understand, with the portability being stable, we could >>>> >>>> theoretically upgrade the SDK without upgrading the runtime components= . >>>> >>>> That would allow us to defer the upgrade for a longer time. >>>> >>>> >>>> Best, >>>> >>>> Max >>>> >>>> >>>> --0000000000008f87a50575b9c466 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable
Hi guys,

Isn= t the issue "only" that beam has this code instead of engines?

Assuming beam runner facin= g api is stable - which must be the case anyway - and that each engine has = its integration (flink-beam instead of beam-runners-flink), then this issue= disappears by construction.

It also has the advantage to have a better maintenance.

Side note: this is what happent which= arquillian, originally the community did all adapters impl then each vendo= r took it back in house to make it better.

Any way to work in that direction maybe?

=
Le jeu. 13 sept. 2018 00:49, Th= omas Weise <thw@apache.org> a = =C3=A9crit=C2=A0:
= The main problem here is that users are forced to upgrade infrastructure to= obtain new features in Beam, even when those features actually don't r= equire such changes. As an example, another update to Flink 1.6.0 was propo= sed (without supporting new functionality in Beam) and we already know that= it breaks compatibility (again).

I think that upgrading= to a Flink X.Y.0 version isn't a good idea to start with. But besides = that, if we want to grow adoption, then we need to focus on stability and d= elivering improvements to Beam without disrupting users.

In the specific case, ideally the surface of Flink would be backward= compatible, allowing us to stick to a minimum version and be able to submi= t pipelines to Flink endpoints of higher versions. Some work in that direct= ion is underway (like versioning the REST API). FYI, lowest common version = is what most projects that depend on Hadoop 2.x follow.

Since Beam with Flink 1.5.x client won't talk to Flink 1.6 and th= ere are code changes required to make it compile, we would need to come up = with a more involved strategy to support multiple Flink versions. Till then= , I would prefer we favor existing users over short lived experiments, whic= h would mean stick with 1.5.x and not support 1.6.0.

Thanks,
Thomas


On Wed, Sep 12, 2018 at 1:15 PM Lukasz Cwik <= lc= wik@google.com> wrote:
As others have already suggested, I also believe LTS releases i= s the best we can do as a community right now until portability allows us t= o decouple what a user writes with and how it runs (the SDK and the SDK env= ironment) from the runner (job service=C2=A0+ shared common runner libs=C2= =A0+ Flink/Spark/Dataflow/Apex/Samza/...).

Dataflow woul= d be highly invested in having the appropriate tooling within Apache Beam t= o support multiple SDK versions against a runner. This in turn would allow = people to use any SDK with any runner and as Robert had mentioned, certain = optimizations and features would be disabled depending on the capabilities = of the runner and the capabilities of the SDK.

=


On Wed, Sep 12, 2018 at 6:38 AM Robert Bradshaw <robertwb@g= oogle.com> wrote:
The target= audience is people who want to use the latest Beam but do not want to use = the latest=C2=A0version of the runner, right?=C2=A0

I think this will be somewhat (though not entirely) addressed by Beam LTS= releases, where those not wanting to upgrade the runner at least have a we= ll-supported version of Beam. In the long term, we have the division
<= div>
=C2=A0 =C2=A0 Runner <-> BeamRunnerSpecificCode &l= t;-> CommonBeamRunnerLibs <-> SDK.

(which= applies to the job submission as well as execution).

<= div>Insomuch as the BeamRunnerSpecificCode uses the public APIs of the runn= er, hopefully upgrading the runner for minor versions should be a no-op, an= d we can target the lowest version of the runner that makes sense, allowing= the user to link against higher versions at his or her discretion. We shou= ld provide built targets that allow this. For major versions, it may make s= ense to have two distinct=C2=A0BeamRunnerSpecificCode libraries (which may = or may not share some common code). I hope these wrappers are not too thick= .=C2=A0

There is a tight coupling at the=C2=A0Beam= RunnerSpecificCode <-> CommonBeamRunnerLibs layer, but hopefully the = bulk of the code lives on the right hand side and can be updated as needed = independent of the runner. There may be code of the form "if the runne= r supports X, do this fast path, otherwise, do this slow path (or reject th= e pipeline).=C2=A0

I hope the=C2=A0CommonBeamRunne= rLibs <-> SDK coupling is fairly loose, to the point that one could u= se SDKs from different versions of Beam (or even developed outside of Beam)= with an older/newer runner. We may need to add versioning to the Fn/Runner= /Job API itself to support this. Right now of course we're still in a p= re-1.0, rapid-development phase wrt this API.=C2=A0




On Wed, Sep 12, 2018 at 2:10 PM Etienne Chauchot &l= t;echauchot@apache.org> wrote:
Hi Max,

I totally agree with your p= oints especially the users priorities (stick to the already working version= ) , and the need to leverage important new features. It is indeed a diffic= ult balance to find .

I can talk for a part I know= : for the Spark runner, the aim was to support Dataset native spark API (in= place of RDD). For that we needed to upgrade to spark 2.x (and we will pro= bably leverage Beam Row as well).
But such an upgrade is a good a= mount of work which makes it difficult to commit on a schedule such as &quo= t;if there is a major new feature on an execution engine that we want to le= verage, then the upgrade in Beam will be done within x months".
<= div>
Regarding your point on portability : decoupling SDK fro= m runner with runner harness and SDK harness might make pipeline authors wo= rk easy regarding pipeline maintenance. But, still, if we upgrade runner li= bs, then the users might have their runner harness not work with their engi= ne version.
If such SDK/runner decoupling is 100% functional, th= en we could imaging having multiple runner harnesses shipping different ver= sions of the runner libs to solve this problem.
But we would need= to support more than one version of the runner libs. We chose not to do th= is on spark runner.

WDYT ?

Best
Etienne


Le mardi 1= 1 septembre 2018 =C3=A0 15:42 +0200, Maximilian Michels a =C3=A9crit=C2=A0:= 
Hi Beamers,

I=
n the light of the discussion about Beam LTS releases, I'd like to kick=
 
off a thread about how often we upgrade the execution engine of=
 each 
Runner. By upgrade, I mean major/minor versions which typi=
cally break 
the binary compatibility of Beam pipelines.

For the Flink Runner, we try to track the latest stable ve=
rsion. Some 
users reported that this can be problematic, as it r=
equires them to 
potentially upgrade their Flink cluster with a n=
ew version of Beam.

 From a developer's perspe=
ctive, it makes sense to migrate as early as 
possible to the new=
est version of the execution engine, e.g. to leverage 
the newest=
 features. From a user's perspective, you don't care about the 
latest features if your use case still works with Beam.
We have to please both parties. So I'd suggest to upgrade =
the execution 
engine whenever necessary (e.g. critical new featu=
res, end of life of 
current version). On the other hand, the upc=
oming Beam LTS releases will 
contain a longer-supported version.=


Maybe we don't need to discuss much about thi=
s but I wanted to hear what 
the community has to say about it. P=
articularly, I'd be interested in 
how the other Runner autho=
rs intend to do it.

As far as I understand, with t=
he portability being stable, we could 
theoretically upgrade the =
SDK without upgrading the runtime components. 
That would allow u=
s to defer the upgrade for a longer time.

Best,
Max

--0000000000008f87a50575b9c466--