Return-Path: X-Original-To: apmail-couchdb-user-archive@www.apache.org Delivered-To: apmail-couchdb-user-archive@www.apache.org Received: from mail.apache.org (hermes.apache.org [140.211.11.3]) by minotaur.apache.org (Postfix) with SMTP id E6508186BA for ; Thu, 1 Oct 2015 20:06:45 +0000 (UTC) Received: (qmail 32634 invoked by uid 500); 1 Oct 2015 20:06:44 -0000 Delivered-To: apmail-couchdb-user-archive@couchdb.apache.org Received: (qmail 32568 invoked by uid 500); 1 Oct 2015 20:06:44 -0000 Mailing-List: contact user-help@couchdb.apache.org; run by ezmlm Precedence: bulk List-Help: List-Unsubscribe: List-Post: List-Id: Reply-To: user@couchdb.apache.org Delivered-To: mailing list user@couchdb.apache.org Received: (qmail 32555 invoked by uid 99); 1 Oct 2015 20:06:44 -0000 Received: from Unknown (HELO spamd2-us-west.apache.org) (209.188.14.142) by apache.org (qpsmtpd/0.29) with ESMTP; Thu, 01 Oct 2015 20:06:44 +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 E15C21A3516 for ; Thu, 1 Oct 2015 20:06:43 +0000 (UTC) X-Virus-Scanned: Debian amavisd-new at spamd2-us-west.apache.org X-Spam-Flag: NO X-Spam-Score: 2.901 X-Spam-Level: ** X-Spam-Status: No, score=2.901 tagged_above=-999 required=6.31 tests=[DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, HTML_MESSAGE=3, RCVD_IN_DNSWL_BLOCKED=0.001, 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=googlemail.com Received: from mx1-us-east.apache.org ([10.40.0.8]) by localhost (spamd2-us-west.apache.org [10.40.0.9]) (amavisd-new, port 10024) with ESMTP id KSJtlB7amnip for ; Thu, 1 Oct 2015 20:06:37 +0000 (UTC) Received: from mail-vk0-f47.google.com (mail-vk0-f47.google.com [209.85.213.47]) by mx1-us-east.apache.org (ASF Mail Server at mx1-us-east.apache.org) with ESMTPS id DA8E942B5D for ; Thu, 1 Oct 2015 20:06:36 +0000 (UTC) Received: by vkao3 with SMTP id o3so47570768vka.2 for ; Thu, 01 Oct 2015 13:06:30 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=googlemail.com; s=20120113; h=mime-version:in-reply-to:references:date:message-id:subject:from:to :content-type; bh=iNh8nlPx2v79XQs9CUk7AIMeTa7zJUGvkEV9etJH4hk=; b=R54Yy9JgOcdbM1Zjum8seRDZKJmxsV0sOVNziJUu3R6CQrW75ftzfM3fG8ew8AX1WY y9SSB9Kui+CHIMUPLd6klZacHyrIvPtkO5Me7sNrjYmIo2v3zl4benS45POGCFFFn5L+ 86M7dU/a0W0tus/dE2kfU8pPZmWWJwbNExl/GIBx12ru0m64Yd+JgC47vl/xXMxaBNxD SsRu1HxIozkaD8VpZMNsrD1k2twcJSSWuRSYg/8UjOfllfWFs/5hofv3CKcHwdtox00+ JdEYRkV2mVTo3fDSWRNvHQF61oxL+JryraxmDBojNOBUCfPLRGb/jRtkI/5kbnUc3P7M KU1Q== MIME-Version: 1.0 X-Received: by 10.31.5.205 with SMTP id 196mr8212966vkf.88.1443729990349; Thu, 01 Oct 2015 13:06:30 -0700 (PDT) Received: by 10.31.160.85 with HTTP; Thu, 1 Oct 2015 13:06:30 -0700 (PDT) In-Reply-To: <2E5B7DDB76C6D547B3791C5FCD3A35E308B993E9@SJCPEX01CL02.citrite.net> References: <2E5B7DDB76C6D547B3791C5FCD3A35E308B9556D@SJCPEX01CL02.citrite.net> <2E5B7DDB76C6D547B3791C5FCD3A35E308B95A28@SJCPEX01CL02.citrite.net> <2E5B7DDB76C6D547B3791C5FCD3A35E308B97B38@SJCPEX01CL02.citrite.net> <2E5B7DDB76C6D547B3791C5FCD3A35E308B98FA1@SJCPEX01CL02.citrite.net> <2E5B7DDB76C6D547B3791C5FCD3A35E308B993E9@SJCPEX01CL02.citrite.net> Date: Thu, 1 Oct 2015 22:06:30 +0200 Message-ID: Subject: Re: stuck indexer process causes view to stop working From: Sebastian Rothbucher To: "user@couchdb.apache.org" Content-Type: multipart/alternative; boundary=001a1143dda039feec05211095a3 --001a1143dda039feec05211095a3 Content-Type: text/plain; charset=UTF-8 Hi Chris, when you have docs that are significantly different from the others (x-tra complex / large) maybe it helps 2 leave them out. Another idea: try tee-ing out what goes back and forth between the couch main process and the indexer process (might become VERY large, but maybe the very end provides some insight). Hope this helps a little - let us know Best Sebastian On Thu, Oct 1, 2015 at 8:01 PM, Chris Thro wrote: > So it just happened again and in the logs I see the following: > [Thu, 01 Oct 2015 17:55:01 GMT] [info] [<0.191.0>] Starting index update > for db: desktopnotification idx: _design/deskto > pnotification > [Thu, 01 Oct 2015 17:55:01 GMT] [info] [<0.191.0>] Index update finished > for db: desktopnotification idx: _design/deskto > Pnotification > > The indexer through shows it started on 2015-10-1 10:55:03 and has made no > progress. > 17:55:01 GMT is 10:55:01 PST > > We are running these on aws. > > -----Original Message----- > From: Chris Thro [mailto:Chris.Thro@citrix.com] > Sent: Thursday, October 01, 2015 9:42 AM > To: user@couchdb.apache.org > Subject: RE: stuck indexer process causes view to stop working > > We are on 1.6.1 > I think this is the erland version R14B04 We are unable to produce the > problem at will. Can you clarify what would constitute a minimal > environment? > > -----Original Message----- > From: Alexander Shorin [mailto:kxepal@gmail.com] > Sent: Thursday, October 01, 2015 8:57 AM > To: user@couchdb.apache.org > Subject: Re: stuck indexer process causes view to stop working > > Hi, > > I'm not sure, but I think the root of evil is somewhere near here: > > [Mon, 28 Sep 2015 12:11:31 GMT] [debug] [<0.86.0>] New task status for > <0.13964.26>: [{changes_done,2}, > {database, > > <<"desktopnotification">>}, > {design_document, > > <<"_design/desktopnotification">>}, > {progress,200}, > > {started_on,1443442291}, > {total_changes,1}, > {type,indexer}, > > {updated_on,1443442291}] > > > Note that progress is 200, while it cannot go beyond 100. So it seems like > there is a bug. At least this looks quite suspicious. > > Few more questions: > - What's your CouchDB and Erlang versions? > - Can you try to upgrade to latest CouchDB release 1.6.1 and see if it > still happens? > - Can you provide minimal environment to reproduce the issue? > > -- > ,,,^..^,,, > > > On Thu, Oct 1, 2015 at 6:51 PM, Chris Thro wrote: > > It happened again on the same nodes. There is still nothing in the > logs. Are there any bug related to stuck indexer? > > > > -----Original Message----- > > From: Chris Thro [mailto:Chris.Thro@citrix.com] > > Sent: Wednesday, September 30, 2015 3:45 PM > > To: user@couchdb.apache.org > > Subject: RE: stuck indexer process causes view to stop working > > > > Thank you for the follow up. I have checked the logs and I do see an > errors. > > I only see the following at the time the indexer got stuck (but this > pops up a lot in the logs) [Mon, 28 Sep 2015 12:11:31 GMT] [debug] > [<0.19987.0>] OS Process #Port<0.8565> Output :: true [Mon, 28 Sep 2015 > 12:11:31 GMT] [debug] [<0.86.0>] New task status for <0.13964.26>: > [{changes_done,2}, > > {database, > > > <<"desktopnotification">>}, > > {design_document, > > > <<"_design/desktopnotification">>}, > > {progress,200}, > > > {started_on,1443442291}, > > {total_changes,1}, > > {type,indexer}, > > > > {updated_on,1443442291}] > > > > > > One other thing I forgot to mention. This happened on two separate > hosts that were setup to replicate between each other. > > > > -----Original Message----- > > From: Alexander Shorin [mailto:kxepal@gmail.com] > > Sent: Wednesday, September 30, 2015 2:17 PM > > To: user@couchdb.apache.org > > Subject: Re: stuck indexer process causes view to stop working > > > > Hi, > > > > Have you check the logs for details? If there was an error that > prevented indexer correct work, it have to be there. > > > > -- > > ,,,^..^,,, > > > > On Thu, Oct 1, 2015 at 12:14 AM, Chris Thro > wrote: > > > >> Hello, > >> > >> I had an issue the other day where if you ran a GET against a view it > >> would hang indefinitely and if you viewed the view in futon it showed > >> no rows and put up a dialog box saying error no reponse. On the > >> status page there was an indexer running on _*design/ that > >> started at 9/28 5:11am and the last update said 9/28 5:11am and the > >> progress was 0 out of 9 changes (0%) this was at least 5 hours past the > start time of the process. > >> I ran killall couchjs as there were a lot of couchjs processes, but > >> they just respawaned and nothing changed. I eventually had to > >> restart the couchdb and that finally removed the indexer process from > >> the status page and allowed the view to work. Do you have any idea > >> what might have caused this issue? Has anyone seen something similar > before? > >> > >> > >> > >> Thank you > >> > >> > >> > >> *Chris Thro* > >> > >> > >> Senior Database Administrator, Operations > >> > >> *T*: +1 805-690-7925 | *M*: +1 707-364-0682 chris.thro@citrix.com > >> > >> > >> > >> > >> > >> > >> > --001a1143dda039feec05211095a3--