Return-Path: <user-return-1082-archive-asf-public=cust-asf.ponee.io@helix.apache.org>
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 DF4A1200C08
	for <archive-asf-public-internal@cust-asf2.ponee.io>; Thu, 26 Jan 2017 19:43:23 +0100 (CET)
Received: by cust-asf.ponee.io (Postfix)
	id DDEE3160B31; Thu, 26 Jan 2017 18:43:23 +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 148B5160B4C
	for <archive-asf-public@cust-asf.ponee.io>; Thu, 26 Jan 2017 19:43:21 +0100 (CET)
Received: (qmail 86371 invoked by uid 500); 26 Jan 2017 18:43:21 -0000
Mailing-List: contact user-help@helix.apache.org; run by ezmlm
Precedence: bulk
List-Help: <mailto:user-help@helix.apache.org>
List-Unsubscribe: <mailto:user-unsubscribe@helix.apache.org>
List-Post: <mailto:user@helix.apache.org>
List-Id: <user.helix.apache.org>
Reply-To: user@helix.apache.org
Delivered-To: mailing list user@helix.apache.org
Received: (qmail 86204 invoked by uid 99); 26 Jan 2017 18:43:20 -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, 26 Jan 2017 18:43:20 +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 42D8F1A06D0
	for <user@helix.apache.org>; Thu, 26 Jan 2017 18:43:20 +0000 (UTC)
X-Virus-Scanned: Debian amavisd-new at spamd2-us-west.apache.org
X-Spam-Flag: NO
X-Spam-Score: 3.58
X-Spam-Level: ***
X-Spam-Status: No, score=3.58 tagged_above=-999 required=6.31
	tests=[DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1,
	HTML_MESSAGE=2, KAM_LINEPADDING=1.2, RCVD_IN_DNSWL_NONE=-0.0001,
	RCVD_IN_MSPIKE_H3=-0.01, RCVD_IN_MSPIKE_WL=-0.01,
	RCVD_IN_SORBS_SPAM=0.5, SPF_PASS=-0.001, URIBL_BLOCKED=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-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 t6_TNUTBCFZo for <user@helix.apache.org>;
	Thu, 26 Jan 2017 18:43:13 +0000 (UTC)
Received: from mail-ot0-f177.google.com (mail-ot0-f177.google.com [74.125.82.177])
	by mx1-lw-us.apache.org (ASF Mail Server at mx1-lw-us.apache.org) with ESMTPS id 0E1985F46F
	for <user@helix.apache.org>; Thu, 26 Jan 2017 18:43:13 +0000 (UTC)
Received: by mail-ot0-f177.google.com with SMTP id 32so22752372oth.3
        for <user@helix.apache.org>; Thu, 26 Jan 2017 10:43:13 -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
         :cc;
        bh=+gh61ClZNFIh3ckCJ1OaoLRlpo+c3luDqR+VHWdkJ6M=;
        b=M1oPsNXyNO6ubp5aMTwXovWjLoSPomU2p0BH5xm1XykU2CLHWyMkJOVO11/o5klm1i
         nHC5GpLXvnOQt7aPwDO4b9QD0V3IaMaKqZLs5YKsJjHZ9OIR7DOush68VG9fI7lR8T+I
         8p2LjT14Zr6koxANmsX2raj8I2u1Idz1PiNxRdE+9OJjMV5q9p740z7+IOnpGrpuGHEJ
         aUhm+/gfWM+MxYPo3jJwwmhB6riWV1GkznAH4WXtVu4ZHj0sFYShhvozRAQd+qihwyVu
         aU+fCFJw6ovJCTROs+HSZpGumGA+Rzy6qpi0YGVM/CGRm6XADxELBXRuHQbpEDnGDey5
         2f3Q==
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:cc;
        bh=+gh61ClZNFIh3ckCJ1OaoLRlpo+c3luDqR+VHWdkJ6M=;
        b=WsrzxmoTLPJC3aeJDnMjAUj/HP92d0tt+35yMISap+H8rWlS/KvuEwvKcadMKmyX/b
         N6KmhTKhMW2pPE/DFTyKelIq7RpfdmNmgm6/AR4XRocnJh+bhuO7iquH/S/+40di8JWu
         /+oFtJr9duZXZSKZdn72Bys19eUh3OiR6l4iuqUG7gX0FcZC6P4AH/7hjL8uuG3ZzqRY
         ZxBhV/ziQlVLqV7Cmn00Bzk4kkRZDpXkyuuDr/zJnknmEgxTR2xApPbYDp1r7+c6eYA/
         yiFtqy4BiGjCx2zzrwxm05RAomQn4iR6IN4DdJIe8Iy6ViMnmPHpfAeqIIz2wJkl/JVe
         iS5A==
X-Gm-Message-State: AIkVDXIdQI/+bJNpDKCQIIRTFErPw4M+24mx8ksnHQ4yDnNDgGQn//SV3gxYwXxEj8WGt0cNHW59qscn2+3O2A==
X-Received: by 10.157.38.162 with SMTP id l31mr1760859otb.58.1485456186310;
 Thu, 26 Jan 2017 10:43:06 -0800 (PST)
MIME-Version: 1.0
Received: by 10.182.108.74 with HTTP; Thu, 26 Jan 2017 10:43:05 -0800 (PST)
In-Reply-To: <601506644E25B342B1ACD8B1EC5D43CFA332BFC7@CAEXE01.integral.com>
References: <601506644E25B342B1ACD8B1EC5D43CFA332BE76@CAEXE01.integral.com>
 <CABaj-QaHwrwm7sjXX5DrMO=eN4C+5goJVXJHkvqhKGT4piWA+A@mail.gmail.com>
 <601506644E25B342B1ACD8B1EC5D43CFA332BED9@CAEXE01.integral.com>
 <CABaj-QZWR8bghWijvnrY3qLu=wCmjc3NYv72tRRB0VA+JJK2zA@mail.gmail.com> <601506644E25B342B1ACD8B1EC5D43CFA332BFC7@CAEXE01.integral.com>
From: kishore g <g.kishore@gmail.com>
Date: Thu, 26 Jan 2017 10:43:05 -0800
Message-ID: <CABaj-QZwZrLXiAL0OU_8efEJoK7CJvy3JPz_f_p1oNSXSaAfqw@mail.gmail.com>
Subject: Re: Double assignment , when participant is not able to establish
 connection with zookeeper quorum
To: "user@helix.apache.org" <user@helix.apache.org>
Cc: "dev@helix.incubator.apache.org" <dev@helix.incubator.apache.org>
Content-Type: multipart/related; boundary=001a1143921850b9c8054703b884
archived-at: Thu, 26 Jan 2017 18:43:24 -0000

--001a1143921850b9c8054703b884
Content-Type: multipart/alternative; boundary=001a1143921850b9c5054703b883

--001a1143921850b9c5054703b883
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Can you please file a ticket and we can probably add this in the next
release.

Minor typo in my email.


   1. There is a pathological case where all zookeeper nodes get
   partitioned/crash/GC. In this case, we will make all participants
   disconnect and assume they don't own the partition. But when zookeepers
   come out of GC, it can continue as if nothing happened i.e it does not
   account for the time when its down. I can't think of a good solution for
   this scenario. Moreover, we *cannot* differentiate between a participant
   GC'ing/partitioned v/s ZK ensemble crash/partition/GC. This is typically
   avoided by ensuring ZK servers are deployed on different racks.


On Thu, Jan 26, 2017 at 10:34 AM, Subramanian Raghunathan <
subramanian.raghunathan@integral.com> wrote:

> Totally concur your thought,  a config based approach would be better .It
> could be tuned based on the acceptable tolerance and consistency.
>
>
>
> Thanks,
>
> Subramanian.
>
>
>
> Tel: +1 (650) 424 4655 <(650)%20424-4655>
>
>
>
> 3400 Hillview Ave, Building 4
>
> Palo Alto, CA 94304
>
> www.integral.com
>
> [image: Logo_signature_block]
> <http://www.integral.com/fxcloud_features/risk_management.html#ym>
>
> NOTICE: This e-mail message and any attachments, which may contain
> confidential information, are to be viewed solely by the intended recipie=
nt
> of Integral Development Corp. For further information, please visit
> http://www.integral.com/about/disclaimer.html.
>
>
>
>
>
>
>
> *From:* kishore g [mailto:g.kishore@gmail.com]
> *Sent:* Wednesday, January 25, 2017 7:12 PM
>
> *To:* user@helix.apache.org
> *Cc:* dev@helix.incubator.apache.org
> *Subject:* Re: Double assignment , when participant is not able to
> establish connection with zookeeper quorum
>
>
>
> Helix can handle this and probably should. Couple of challenges here are
>
>    1. How to generalize this across all use cases. This is a
>    trade-off between availability and ensuring there is only one leader p=
er
>    partition.
>    2. There is a pathological case where all zookeeper nodes get
>    partitioned/crash/GC. In this case, we will make all participants
>    disconnect and assume they don't own the partition. But when zookeeper=
s
>    come out of GC, it can continue as if nothing happened i.e it does not
>    account for the time when its down. I can't think of a good solution f=
or
>    this scenario. Moreover, we can differentiate between a participant
>    GC'ing/partitioned v/s ZK ensemble crash/partition/GC. This is typical=
ly
>    avoided by ensuring ZK servers are deployed on different racks.
>
> Having said that, I think implementing a config based solution is worth
> it.
>
>
>
>
>
>
>
>
>
>
>
> On Wed, Jan 25, 2017 at 4:57 PM, Subramanian Raghunathan <
> subramanian.raghunathan@integral.com> wrote:
>
> Hi Kishore ,
>
>
>
>                 Thank you for the confirmation , yes we had solved it in
> similar lines and it did work for us (listening on the disconnect event
> from ZK).
>
>
>
>                 From the double assignment point of view is it an expecte=
d
> behavior from Helix and the users to handle the same ? Is there any plans
> to fix the same in future ?
>
>
>
> Because what I had observed when the network is flapping helix does handl=
e
> it by calling reset () for the partition(s) from the (disconnect()), then
> why not in this case ?
>
>
>
> void org.apache.helix.manager.zk.ZkHelixConnection.handleStateChanged(Kee=
perState
> state) throws Exception
>
>
>
> if (isFlapping()) {
>
>         LOG.error("helix-connection: " + this + ", sessionId: " +
> _sessionId
>
>             + " is flapping. diconnect it. " + " maxDisconnectThreshold: =
"
>
>             + _maxDisconnectThreshold + " disconnects in " +
> _flappingTimeWindowMs + "ms");
>
>         disconnect();
>
>       }
>
>
>
>
>
> Thanks & Regards,
>
> Subramanian.
>
>
>
> Tel: +1 (650) 424 4655 <(650)%20424-4655>
>
>
>
> 3400 Hillview Ave, Building 4
>
> Palo Alto, CA 94304
>
> www.integral.com
>
> [image: Logo_signature_block]
> <http://www.integral.com/fxcloud_features/risk_management.html#ym>
>
> NOTICE: This e-mail message and any attachments, which may contain
> confidential information, are to be viewed solely by the intended recipie=
nt
> of Integral Development Corp. For further information, please visit
> http://www.integral.com/about/disclaimer.html.
>
>
>
>
>
>
>
> *From:* kishore g [mailto:g.kishore@gmail.com]
> *Sent:* Wednesday, January 25, 2017 4:45 PM
> *To:* user@helix.apache.org
> *Cc:* dev@helix.incubator.apache.org
> *Subject:* Re: Double assignment , when participant is not able to
> establish connection with zookeeper quorum
>
>
>
> After few seconds, the participant N1 gets a disconnect event from ZK. At
> this time, schedule a timer task for (30  - X) seconds. 30 is the session
> timeout and X can vary from 0 to 30 depending on how long you are ok to n=
ot
> have a P1 being down.
>
>
>
> When the timer task kicks in and N1 is still disconnected from the
> cluster, assume that this N1 is no longer the owner of P1.
>
>
>
> After 30 seconds, Helix will notice that N1 is network partitioned and
> will assign P1 to N2.
>
> This will ensure that there is no overlap.
>
>
>
> Will that work for you?
>
>
>
>
>
> On Wed, Jan 25, 2017 at 4:17 PM, Subramanian Raghunathan <
> subramanian.raghunathan@integral.com> wrote:
>
> Hi ,
>
>
>
> Double assignment , when participant is not able to establish connection
> with zookeeper quorum
>
>
>
> Following is the  set up.
>
>
>
> Version(s) :
>
>                                 Helix: 0.7.1
>
>                                 Zookeeper:3.3.4
>
>
>
> - State Model: OnlineOffline
>
> - Controller (leader elected from one of the cluster nodes)
>
> - Single resources with partitions.
>
> - Full auto rebalancer
>
>
>
> -Zookeeper quorum (3 nodes)
>
>
>
> When one participant loses the zookeeper connection (It=E2=80=99s not abl=
e to
> connect to any of the zookeepers , a typical occurrence we faced was swit=
ch
> failure from that rack)
>
>
>
>   ---- >  The partition (P1) for which this participant (say Node N1) is
> online is still maintained
>
>
>
> Meanwhile since it loses the ephemeral  node in zookeeper , the rebalance=
r
> gets triggered and it reallocates the partition (P1) to another participa=
nt
> node (say Node N2) to become online  @ time T1
>
>
>
>                 ---- >  *After this both N1 and N2 are acting as online
> for the same Partition (P1) *
>
>
>
> But as soon as participant in (say Node N1) is able to re-establish the
> zookeeper connection  @ time T2
>
>                 ---- >  Reset gets called on the partition in participant
> (say Node N1)
>
>
>
> Double assignment:
>
> The question here is this an expected behavior that both nodes N1 and N2
> could be online for the same Partition (P1) between time (T1-T2) ? Any
> responses on the same would be appreciated.
>
>
>
> Thanks & Regards,
>
> Subramanian.
>
>
>
> 3400 Hillview Ave, Building 4
>
> Palo Alto, CA 94304
>
> www.integral.com
>
> [image: Logo_signature_block]
> <http://www.integral.com/fxcloud_features/risk_management.html#ym>
>
> NOTICE: This e-mail message and any attachments, which may contain
> confidential information, are to be viewed solely by the intended recipie=
nt
> of Integral Development Corp. For further information, please visit
> http://www.integral.com/about/disclaimer.html.
>
>
>
>
>
>
>
>
>
>
>

--001a1143921850b9c5054703b883
Content-Type: text/html; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

<div dir=3D"ltr"><div><span style=3D"font-size:12.8px">Can you please file =
a ticket and we can probably add this in the next release.=C2=A0</span></di=
v><div><span style=3D"font-size:12.8px"><br></span></div><div><span style=
=3D"font-size:12.8px">Minor typo in my email.</span></div><div><span style=
=3D"font-size:12.8px"><br></span></div><ol style=3D"font-size:12.8px"><li s=
tyle=3D"margin-left:15px">There is a pathological case where all zookeeper =
nodes get partitioned/crash/GC. In this case, we will make all participants=
 disconnect and assume they don&#39;t own the partition. But when zookeeper=
s come out of GC, it can continue as if nothing happened i.e it does not ac=
count for the time when its down. I can&#39;t think of a good solution for =
this scenario. Moreover, we <b style=3D"background-color:rgb(106,168,79)">c=
annot</b> differentiate between a participant GC&#39;ing/partitioned v/s ZK=
 ensemble crash/partition/GC. This is typically avoided by ensuring ZK serv=
ers are deployed on different racks.</li></ol></div><div class=3D"gmail_ext=
ra"><br><div class=3D"gmail_quote">On Thu, Jan 26, 2017 at 10:34 AM, Subram=
anian Raghunathan <span dir=3D"ltr">&lt;<a href=3D"mailto:subramanian.raghu=
nathan@integral.com" target=3D"_blank">subramanian.raghunathan@integral.com=
</a>&gt;</span> wrote:<br><blockquote class=3D"gmail_quote" style=3D"margin=
:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">





<div lang=3D"EN-US" link=3D"blue" vlink=3D"purple">
<div class=3D"m_5847791235137943666WordSection1">
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif">Totally concur your thought, =C2=A0a config based a=
pproach would be better .It could be tuned based on the acceptable toleranc=
e and consistency.<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif"><u></u>=C2=A0<u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif;color:#767171">Thanks,<u></u><u></u></span></p><span=
 class=3D"">
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif;color:#767171">Subramanian.<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif;color:#767171"><u></u>=C2=A0<u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif;color:#767171">Tel: <a href=3D"tel:(650)%20424-4655"=
 value=3D"+16504244655" target=3D"_blank">+1 (650) 424 4655</a></span><span=
 style=3D"font-size:10.0pt;font-family:&quot;Calibri&quot;,sans-serif;color=
:gray">
<u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif;color:#767171"><u></u>=C2=A0<u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif;color:#767171">3400 Hillview Ave, Building 4<u></u><=
u></u></span></p>
<p class=3D"MsoNormal" style=3D"margin-bottom:12.0pt"><span style=3D"font-s=
ize:10.0pt;font-family:&quot;Calibri&quot;,sans-serif;color:#767171">Palo A=
lto, CA 94304</span><span style=3D"font-size:6.0pt;font-family:&quot;Calibr=
i&quot;,sans-serif;color:#767171"><u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif;color:#0070c0"><a href=3D"http://www.integral.com/" =
title=3D"mailto:dennis.wong@integral.com" target=3D"_blank"><span style=3D"=
color:#0070c0">www.integral.com</span></a></span><span style=3D"font-size:1=
0.0pt;font-family:&quot;Calibri&quot;,sans-serif;color:#1f497d"><u></u><u><=
/u></span></p>
<p class=3D"MsoNormal"><a href=3D"http://www.integral.com/fxcloud_features/=
risk_management.html#ym" title=3D"Winner: FXWeek e-FX Awards 2016" target=
=3D"_blank"><span style=3D"font-size:10.0pt;font-family:&quot;Calibri&quot;=
,sans-serif;text-decoration:none"><img border=3D"0" width=3D"290" height=3D=
"120" style=3D"width:3.0208in;height:1.25in" id=3D"m_5847791235137943666Pic=
ture_x0020_3" src=3D"cid:image001.png@01D277BF.C647AA50" alt=3D"Logo_signat=
ure_block"></span></a><span style=3D"font-size:10.0pt;font-family:&quot;Cal=
ibri&quot;,sans-serif;color:gray"><br>
</span><span style=3D"font-size:10.0pt;font-family:&quot;Calibri&quot;,sans=
-serif;color:#767171"><br>
NOTICE: This e-mail message and any attachments, which may contain confiden=
tial information, are to be viewed solely by the intended recipient of Inte=
gral Development Corp. For further information, please visit
</span><span style=3D"font-size:10.0pt;font-family:&quot;Calibri&quot;,sans=
-serif;color:#0070c0"><a href=3D"http://www.integral.com/about/disclaimer.h=
tml" target=3D"_blank"><span style=3D"color:#0070c0">http://www.integral.co=
m/about/<wbr>disclaimer.html</span></a></span><span style=3D"font-size:10.0=
pt;font-family:&quot;Calibri&quot;,sans-serif;color:gray">.</span><span sty=
le=3D"font-size:10.0pt;font-family:&quot;Calibri&quot;,sans-serif;color:#1f=
497d">
</span><span style=3D"font-size:10.0pt;font-family:&quot;Calibri&quot;,sans=
-serif;color:gray"><u></u><u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif"><u></u>=C2=A0<u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif"><u></u>=C2=A0<u></u></span></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif"><u></u>=C2=A0<u></u></span></p>
</span><p class=3D"MsoNormal"><b><span style=3D"font-size:11.0pt;font-famil=
y:&quot;Calibri&quot;,sans-serif">From:</span></b><span style=3D"font-size:=
11.0pt;font-family:&quot;Calibri&quot;,sans-serif"> kishore g [mailto:<a hr=
ef=3D"mailto:g.kishore@gmail.com" target=3D"_blank">g.kishore@gmail.com</a>=
]
<br>
<b>Sent:</b> Wednesday, January 25, 2017 7:12 PM</span></p><div><div class=
=3D"h5"><br>
<b>To:</b> <a href=3D"mailto:user@helix.apache.org" target=3D"_blank">user@=
helix.apache.org</a><br>
<b>Cc:</b> <a href=3D"mailto:dev@helix.incubator.apache.org" target=3D"_bla=
nk">dev@helix.incubator.apache.org</a><br>
<b>Subject:</b> Re: Double assignment , when participant is not able to est=
ablish connection with zookeeper quorum<u></u><u></u></div></div><p></p><di=
v><div class=3D"h5">
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<div>
<p class=3D"MsoNormal">Helix can handle this and probably should. Couple of=
 challenges here are<u></u><u></u></p>
<div>
<ol start=3D"1" type=3D"1">
<li class=3D"MsoNormal" style=3D"margin-left:0in">
How to generalize this across all use cases. This is a trade-off=C2=A0betwe=
en availability and ensuring there is only one leader per partition.=C2=A0<=
u></u><u></u></li><li class=3D"MsoNormal" style=3D"margin-left:0in">
There is a pathological case where all zookeeper nodes get partitioned/cras=
h/GC. In this case, we will make all participants disconnect and assume the=
y don&#39;t own the partition. But when zookeepers come out of GC, it can c=
ontinue as if nothing happened i.e it
 does not account for the time when its down. I can&#39;t think of a good s=
olution for this scenario. Moreover, we can differentiate between a partici=
pant GC&#39;ing/partitioned v/s ZK ensemble crash/partition/GC. This is typ=
ically avoided by ensuring ZK servers are
 deployed on different racks.<u></u><u></u></li></ol>
<div>
<p class=3D"MsoNormal">Having said that, I think implementing a config base=
d solution is worth it.=C2=A0<u></u><u></u></p>
</div>
</div>
<div>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
</div>
<div>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
</div>
<div>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
</div>
<div>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
</div>
</div>
<div>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
<div>
<p class=3D"MsoNormal">On Wed, Jan 25, 2017 at 4:57 PM, Subramanian Raghuna=
than &lt;<a href=3D"mailto:subramanian.raghunathan@integral.com" target=3D"=
_blank">subramanian.raghunathan@<wbr>integral.com</a>&gt; wrote:<u></u><u><=
/u></p>
<blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0i=
n 0in 0in 6.0pt;margin-left:4.8pt;margin-right:0in">
<div>
<div>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif">Hi Kishore ,</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif">=C2=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 Thank you for the confirmation , ye=
s we had solved it in similar lines and it did work for us (listening
 on the </span>disconnect event from ZK<span style=3D"font-size:11.0pt;font=
-family:&quot;Calibri&quot;,sans-serif">).</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif">=C2=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 From the double assignment point of=
 view is it an expected behavior from Helix and the users to handle
 the same ? Is there any plans to fix the same in future ? </span><u></u><u=
></u></p>
<p class=3D"MsoNormal" style=3D"text-indent:.5in">
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,sans-serif"=
>=C2=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal" style=3D"text-indent:.5in">
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,sans-serif"=
>Because what I had observed when the network is flapping helix does handle=
 it by calling reset () for the partition(s) from the (disconnect()), then =
why not in this case ?</span><u></u><u></u></p>
<p class=3D"MsoNormal" style=3D"text-indent:.5in">
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,sans-serif"=
>=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0<wbr>=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0
</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif">void org.apache.helix.manager.zk.<wbr>ZkHelixConnec=
tion.<wbr>handleStateChanged(KeeperState state) throws Exception</span><u><=
/u><u></u></p>
<p class=3D"MsoNormal" style=3D"text-indent:.5in">
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,sans-serif"=
>=C2=A0</span><u></u><u></u></p>
<table class=3D"m_5847791235137943666MsoNormalTable" border=3D"0" cellspaci=
ng=3D"0" cellpadding=3D"0" style=3D"margin-left:92.6pt;border-collapse:coll=
apse">
<tbody>
<tr>
<td width=3D"623" valign=3D"top" style=3D"width:467.5pt;border:solid window=
text 1.0pt;padding:0in 5.4pt 0in 5.4pt">
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif">if (isFlapping()) {</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 LOG.erro=
r(&quot;helix-connection: &quot; + this + &quot;, sessionId: &quot; + _sess=
ionId</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A0=C2=A0 + &quot; is flapping. diconnect it. &quot; + &quot; maxDisc=
onnectThreshold: &quot;</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A0=C2=A0 + _maxDisconnectThreshold + &quot; disconnects in &quot; + =
_flappingTimeWindowMs + &quot;ms&quot;);</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 disconne=
ct();</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 }</span><u></u><u></=
u></p>
</td>
</tr>
</tbody>
</table>
<p class=3D"MsoNormal" style=3D"text-indent:.5in">
<span style=3D"font-size:11.0pt;font-family:&quot;Calibri&quot;,sans-serif"=
>=C2=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0
</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif;color:#767171">Thanks &amp; Regards,</span><u></u><u=
></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif;color:#767171">Subramanian.</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif;color:#767171">=C2=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif;color:#767171">Tel:
<a href=3D"tel:(650)%20424-4655" target=3D"_blank">+1 (650) 424 4655</a></s=
pan><span style=3D"font-size:10.0pt;font-family:&quot;Calibri&quot;,sans-se=
rif;color:gray">
</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif;color:#767171">=C2=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif;color:#767171">3400 Hillview Ave, Building 4</span><=
u></u><u></u></p>
<p class=3D"MsoNormal" style=3D"margin-bottom:12.0pt"><span style=3D"font-s=
ize:10.0pt;font-family:&quot;Calibri&quot;,sans-serif;color:#767171">Palo A=
lto, CA 94304</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif;color:#0070c0"><a href=3D"http://www.integral.com/" =
title=3D"mailto:dennis.wong@integral.com" target=3D"_blank"><span style=3D"=
color:#0070c0">www.integral.com</span></a></span><u></u><u></u></p>
<p class=3D"MsoNormal"><a href=3D"http://www.integral.com/fxcloud_features/=
risk_management.html#ym" title=3D"Winner: FXWeek e-FX Awards 2016" target=
=3D"_blank"><span style=3D"font-size:10.0pt;font-family:&quot;Calibri&quot;=
,sans-serif;text-decoration:none"><img border=3D"0" width=3D"290" height=3D=
"120" style=3D"width:3.0208in;height:1.25in" id=3D"m_5847791235137943666m_1=
443031970177955345Picture_x0020_3" src=3D"cid:image001.png@01D277BF.C647AA5=
0" alt=3D"Logo_signature_block"></span></a><span style=3D"font-size:10.0pt;=
font-family:&quot;Calibri&quot;,sans-serif;color:gray"><br>
</span><span style=3D"font-size:10.0pt;font-family:&quot;Calibri&quot;,sans=
-serif;color:#767171"><br>
NOTICE: This e-mail message and any attachments, which may contain confiden=
tial information, are to be viewed solely by the intended recipient of Inte=
gral Development Corp. For further information, please visit
</span><span style=3D"font-size:10.0pt;font-family:&quot;Calibri&quot;,sans=
-serif;color:#0070c0"><a href=3D"http://www.integral.com/about/disclaimer.h=
tml" target=3D"_blank"><span style=3D"color:#0070c0">http://www.integral.co=
m/about/<wbr>disclaimer.html</span></a></span><span style=3D"font-size:10.0=
pt;font-family:&quot;Calibri&quot;,sans-serif;color:gray">.</span><span sty=
le=3D"font-size:10.0pt;font-family:&quot;Calibri&quot;,sans-serif;color:#1f=
497d">
</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif">=C2=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif">=C2=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:11.0pt;font-family:&quot;Ca=
libri&quot;,sans-serif">=C2=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><b><span style=3D"font-size:11.0pt;font-family:&quot=
;Calibri&quot;,sans-serif">From:</span></b><span style=3D"font-size:11.0pt;=
font-family:&quot;Calibri&quot;,sans-serif"> kishore g [mailto:<a href=3D"m=
ailto:g.kishore@gmail.com" target=3D"_blank">g.kishore@gmail.com</a>]
<br>
<b>Sent:</b> Wednesday, January 25, 2017 4:45 PM<br>
<b>To:</b> <a href=3D"mailto:user@helix.apache.org" target=3D"_blank">user@=
helix.apache.org</a><br>
<b>Cc:</b> <a href=3D"mailto:dev@helix.incubator.apache.org" target=3D"_bla=
nk">dev@helix.incubator.apache.org</a><br>
<b>Subject:</b> Re: Double assignment , when participant is not able to est=
ablish connection with zookeeper quorum</span><u></u><u></u></p>
<div>
<div>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<div>
<div>
<p class=3D"MsoNormal">After few seconds, the participant N1 gets a disconn=
ect event from ZK. At this time, schedule a timer task for (30 =C2=A0- X) s=
econds. 30 is the session timeout and X can vary from 0
 to 30 depending on how long you are ok to not have a P1 being down.=C2=A0<=
u></u><u></u></p>
</div>
<div>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
</div>
<div>
<p class=3D"MsoNormal">When the timer task kicks in and N1 is still disconn=
ected from the cluster, assume that this N1 is no longer the owner of P1.<u=
></u><u></u></p>
</div>
<div>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
</div>
<div>
<p class=3D"MsoNormal">After 30 seconds, Helix will notice that N1 is netwo=
rk partitioned and will assign P1 to N2.=C2=A0<u></u><u></u></p>
</div>
<div>
<p class=3D"MsoNormal">This will ensure that there is no overlap.<u></u><u>=
</u></p>
</div>
<div>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
</div>
<div>
<p class=3D"MsoNormal">Will that work for you?<u></u><u></u></p>
</div>
<div>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
</div>
</div>
<div>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<div>
<p class=3D"MsoNormal">On Wed, Jan 25, 2017 at 4:17 PM, Subramanian Raghuna=
than &lt;<a href=3D"mailto:subramanian.raghunathan@integral.com" target=3D"=
_blank">subramanian.raghunathan@<wbr>integral.com</a>&gt; wrote:<u></u><u><=
/u></p>
<blockquote style=3D"border:none;border-left:solid #cccccc 1.0pt;padding:0i=
n 0in 0in 6.0pt;margin-left:4.8pt;margin-top:5.0pt;margin-right:0in;margin-=
bottom:5.0pt">
<div>
<div>
<p class=3D"MsoNormal">Hi ,<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<p class=3D"MsoNormal">Double assignment , when participant is not able to =
establish connection with zookeeper quorum
<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<p class=3D"MsoNormal">Following is the=C2=A0 set up.<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<p class=3D"MsoNormal">Version(s) :<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0<wbr>=C2=A0 Helix: 0.=
7.1<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=
=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0<wbr>=C2=A0 Zookeeper=
:3.3.4<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<p class=3D"MsoNormal">- State Model: OnlineOffline
<u></u><u></u></p>
<p class=3D"MsoNormal">- Controller (leader elected from one of the cluster=
 nodes)<u></u><u></u></p>
<p class=3D"MsoNormal">- Single resources with partitions.<u></u><u></u></p=
>
<p class=3D"MsoNormal">- Full auto rebalancer<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<p class=3D"MsoNormal">-Zookeeper quorum (3 nodes)<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<p class=3D"MsoNormal">When one participant loses the zookeeper connection =
(It=E2=80=99s not able to connect to any of the zookeepers , a typical occu=
rrence we faced was switch failure from that rack)
<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0 ---- &gt;=C2=A0 The partition (P1) for which =
this participant (say Node N1) is online is still maintained<u></u><u></u><=
/p>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<p class=3D"MsoNormal">Meanwhile since it loses the ephemeral=C2=A0 node in=
 zookeeper , the rebalancer gets triggered and it reallocates the partition=
 (P1) to another participant node (say Node N2) to become
 online =C2=A0@ time T1<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 ---- &gt;=C2=A0
<b>After this both N1 and N2 are acting as online for the same Partition (P=
1) </b>
<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<p class=3D"MsoNormal">But as soon as participant in (say Node N1) is able =
to re-establish the zookeeper connection=C2=A0 @ time T2<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 ---- &gt;=C2=A0 Reset gets called o=
n the partition in participant (say Node N1)
<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=
=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0
<u></u><u></u></p>
<p class=3D"MsoNormal">Double assignment:
<u></u><u></u></p>
<p class=3D"MsoNormal">The question here is this an expected behavior that =
both nodes N1 and N2 could be online for the same Partition (P1) between ti=
me (T1-T2) ? Any responses on the same would be appreciated.<u></u><u></u><=
/p>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;color:#767171">Thank=
s &amp; Regards,</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;color:#767171">Subra=
manian.</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;color:#767171">=C2=
=A0</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;color:#767171">3400 =
Hillview Ave, Building 4</span><u></u><u></u></p>
<p class=3D"MsoNormal" style=3D"margin-bottom:12.0pt"><span style=3D"font-s=
ize:10.0pt;color:#767171">Palo Alto, CA 94304</span><u></u><u></u></p>
<p class=3D"MsoNormal"><span style=3D"font-size:10.0pt;color:#0070c0"><a hr=
ef=3D"http://www.integral.com/" title=3D"mailto:dennis.wong@integral.com" t=
arget=3D"_blank"><span style=3D"color:#0070c0">www.integral.com</span></a><=
/span><u></u><u></u></p>
<p class=3D"MsoNormal"><a href=3D"http://www.integral.com/fxcloud_features/=
risk_management.html#ym" title=3D"Winner: FXWeek e-FX Awards 2016" target=
=3D"_blank"><span style=3D"font-size:10.0pt;text-decoration:none"><img bord=
er=3D"0" width=3D"290" height=3D"120" style=3D"width:3.0208in;height:1.25in=
" id=3D"m_5847791235137943666m_1443031970177955345m_-2723532909891785936Pic=
ture_x0020_3" src=3D"cid:image001.png@01D277BF.C647AA50" alt=3D"Logo_signat=
ure_block"></span></a><span style=3D"font-size:10.0pt;color:gray"><br>
</span><span style=3D"font-size:10.0pt;color:#767171"><br>
NOTICE: This e-mail message and any attachments, which may contain confiden=
tial information, are to be viewed solely by the intended recipient of Inte=
gral Development Corp. For further information, please visit
</span><span style=3D"font-size:10.0pt;color:#0070c0"><a href=3D"http://www=
.integral.com/about/disclaimer.html" target=3D"_blank"><span style=3D"color=
:#0070c0">http://www.integral.com/about/<wbr>disclaimer.html</span></a></sp=
an><span style=3D"font-size:10.0pt;color:gray">.</span><span style=3D"font-=
size:10.0pt;color:#1f497d">
</span><u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
</div>
</div>
</blockquote>
</div>
<p class=3D"MsoNormal">=C2=A0<u></u><u></u></p>
</div>
</div>
</div>
</div>
</div>
</blockquote>
</div>
<p class=3D"MsoNormal"><u></u>=C2=A0<u></u></p>
</div>
</div></div></div>
</div>

</blockquote></div><br></div>

--001a1143921850b9c5054703b883--

--001a1143921850b9c8054703b884
Content-Type: image/png; name="image001.png"
Content-Disposition: inline; filename="image001.png"
Content-Transfer-Encoding: base64
Content-ID: <image001.png@01D277BF.C647AA50>
X-Attachment-Id: dca24e2c5d6ccbe4_0.1
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--001a1143921850b9c8054703b884--