lucene-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Mark Miller (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (SOLR-3846) TestReplicationHandler.test always (?) takes many minutes on OS X (lion)
Date Mon, 17 Sep 2012 15:02:08 GMT

    [ https://issues.apache.org/jira/browse/SOLR-3846?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13457069#comment-13457069
] 

Mark Miller commented on SOLR-3846:
-----------------------------------

I'm pretty sure this was affecting one of my tests on linux as well - it wouldn't hang, but
it would randomly oscillate between taking 2 seconds and 2-3 minutes - the only pertinent
stack trace I could get was in a socket read wait. I've been trying to figure out what's going
on for a couple days now. There was no smoking gun getAddr stack trace like this though. Anyway,
I checked and it had some localhost usage. I'm not home to test on my linux box, but that
test is already looking better on my macbook air (last night it was also taking a long time
on the air - this morning, after the change from localhost, it seems to be consistently faster).

                
> TestReplicationHandler.test always (?) takes many minutes on OS X (lion)
> ------------------------------------------------------------------------
>
>                 Key: SOLR-3846
>                 URL: https://issues.apache.org/jira/browse/SOLR-3846
>             Project: Solr
>          Issue Type: Improvement
>          Components: Build
>    Affects Versions: 4.0-BETA, 5.0
>         Environment: OS X (Lion). Apparently (see Yonik's notes) this does NOT happen
on other op systems.
> java version "1.6.0_35"
> Java(TM) SE Runtime Environment (build 1.6.0_35-b10-428-11M3811)
> Java HotSpot(TM) 64-Bit Server VM (build 20.10-b01-428, mixed mode)
> Solr trunk and 4.x from 16-Sep, but it's been happening for a couple of weeks at least.
>            Reporter: Erick Erickson
>             Fix For: 4.0, 5.0
>
>         Attachments: SOLR-3846.patch, SOLR-3846.patch, stacks.txt
>
>
> Here's the seed was using, but this is apparently unnecessary:
> <JUnit4> says ¡Hola! Master seed: 6785BB3284A15298
> _eventually_ it seems to complete, but it takes many minutes, for instance this was reported
once, but I usually lose patience and ctrl-c out:
> {code}
> [junit4:junit4] Completed on J2 in 2449.62s, 1 test
> [junit4:junit4] 
> [junit4:junit4] JVM J0:     1.21 ..   266.67 =   265.47s
> [junit4:junit4] JVM J1:     1.21 ..   238.33 =   237.12s
> [junit4:junit4] JVM J2:     1.21 ..  2538.60 =  2537.39s
> [junit4:junit4] JVM J3:     0.97 ..   267.37 =   266.40s
> [junit4:junit4] Execution time total: 42 minutes 18 seconds
> {code}
> and a lot of lines like:
> HEARTBEAT J2: 2012-09-16T17:38:38, no events in:  187s, approx. at: TestReplicationHandler.test
> Yonik reports that he can make this happen 100% of the time on OS X/Lion, which squares
with my experience as I recall. Yonik also reports...
> On my linux box (built in '09, PhenomII, HDD) the test takes 50-55 sec.
> On my kids old windows box ('08, athlon64, HDD, Win7) the test takes 88-95 sec.
> On my mac it always takes forever, and I see loops of stuff like this:
> {code}
> SEVERE Master at: http://localhost:62803/solr is not available. Index
> fetch failed. Exception:
> org.apache.solr.client.solrj.SolrServerException: Server refused
> connection at: http://localhost:62803/solr
> [junit4:junit4]   2> 52751 T219 C17 UPDATE [collection1] webapp=/solr
> path=/update params={wt=javabin&version=2} {add=[150]} 0 0
> [junit4:junit4]   2> 52755 T219 C17 UPDATE [collection1] webapp=/solr
> path=/update params={wt=javabin&version=2} {add=[151]} 0 0
> [junit4:junit4]   2> 62758 T215 oash.SnapPuller.fetchLatestIndex
> SEVERE Master at: http://localhost:62803/solr is not available. Index
> fetch failed. Exception:
> {code}
> And I'm soooo happy it's not happening to others and just being swept under the rug,
restores my faith. I should have known better ;)
> See the discussion on the dev list labeled "being a good citizen is hard when you can't
successfully run tests" for more context.
> I don't know how much time I'll have to dive in to it but I'll certainly be happy to
test anyone's patch.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@lucene.apache.org
For additional commands, e-mail: dev-help@lucene.apache.org


Mime
View raw message