http://www.couchbase.com/issues This file is an XML representation of an issue en-us 5.2.4 845 26-12-2012 http://www.couchbase.com/issues/browse/MB-5938 Couchbase Server It works fine with old samples (gamesim-sample), but recent archives (beer-sample) contains much more data and usually fail due to timeout. <br/> <br/> Now it takes at least 40 seconds to load documents and create views. MB-5938

UI needs to deal with very slow docloader (alternatively, docloader is too slow)

Bug Blocker Resolved Fixed Public Aleksey Kondratenko Pavel Paulau 2.0-beta-release-notes Wed, 18 Jul 2012 13:17:44 -0500 Fri, 26 Oct 2012 18:14:36 -0500 Mon, 17 Sep 2012 10:14:18 -0500 recent-builds-2.0 2.0-beta-2 ns_server 1 3 Hm. Users are not generally prepared to wait that much. In theory there should be no problem loading thousands of docs per second. I don&#39;t think beers data is mega docs. 1. before loading these docs we have to create new bucket (if it doesn&#39;t exist). <br/> 2. every document is a separate file (for beers we have more then 6000 files in archive). and current approach to read these files is pretty weird. <br/> 3. we also create new ddocs/views and perform queries on new indices. <br/> <br/> if there is any problem with increasing timeout, I can fix point #2. Increasing timeout is simple. But waiting 40 seconds is a bit too much for users to tolerate. <br/> ok, fixing #2 doesn&#39;t help sometimes. <br/> <br/> python client have extremely low throughput (regardless environment). and this is behind cbdocloader. <br/> thus we need ~15-20 seconds to load these sample items. <br/> <br/> I understand that 40 seconds are too long. but nobody is going to fix/improve old sdk branch, which is only compatible with python 2.4. Trying longer timeout. That sucks. We don&#39;t have good feedback for users. Can we just make the load asynchronous and log problems? Integrated in github-ns-server-2-0 #412 (See [<a href="http://qa.hq.northscale.net/job/github-ns-server-2-0/412/">http://qa.hq.northscale.net/job/github-ns-server-2-0/412/</a>])<br/> &nbsp;&nbsp;&nbsp;&nbsp;<a href="http://www.couchbase.com/issues/browse/MB-5938" title="UI needs to deal with very slow docloader (alternatively, docloader is too slow)"><strike>MB-5938</strike></a>: bumped up docloader timeout to 60 seconds (Revision f6d28497ddb2270962b8ba73284da66237120be8)<br/> <br/> &nbsp;&nbsp;&nbsp;&nbsp;&nbsp;Result = SUCCESS<br/> Aliaksey Artamonau : <br/> Files : <br/> * src/menelaus_web.erl<br/> Assigning to Pavel for UI-only attempt to fix I had tried to POST on <a href="http://localhost:9001/sampleBuckets/install">http://localhost:9001/sampleBuckets/install</a> and after 2.8min of pending I got error 400 from server with the following response <br/> {&quot;error&quot;:true,&quot;type&quot;:&quot;docloader_failed&quot;,&quot;reason&quot;:&quot;There was an unexpected error.&quot;} <br/> <br/> In this case SAMPLES_LOADING_TIMEOUT was 12000000. There&#39;s blocker level bug on docloader for that Also look at <br/> <a href="http://www.couchbase.com/issues/browse/MB-6274">http://www.couchbase.com/issues/browse/MB-6274</a> <br/> <br/> for other docloader improvements. Not actually fixed. Sorry. Meant to assign to Pavel <a href="http://review.couchbase.org/20834">http://review.couchbase.org/20834</a> I&#39;ve uploaded alternative commit to gerrit. Sample loading call will now return as soon as buckets are created, then data will be loaded in background and task will be reported via list of tasks (without progress report however) in right top corner. Sample data will now be asynchronously loaded. Loading sample data will be <br/> executed as a background task then later reported in web console <br/> upon completion. Date of First Response Wed, 18 Jul 2012 15:08:50 -0500 Rank 4617 Sprint Priority 2.0 Sprint Status Time In Status