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-4006) Many tests on Apache Jenkins are failing with lingering threads.
Date Sat, 10 Nov 2012 13:05:25 GMT

    [ https://issues.apache.org/jira/browse/SOLR-4006?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13494669#comment-13494669
] 

Mark Miller commented on SOLR-4006:
-----------------------------------

I don't think the spec matters much - I've got two versions in envs with different Comps and
different host hardware and different VM hardware.

I think the key is simply installing freebsd and turning on the hardcore black hole mode,
eg

{noformat}
# sysctl -w net.inet.tcp.blackhole=2
net.inet.tcp.blackhole: 0 -> 2
# sysctl -w net.inet.udp.blackhole=1
net.inet.udp.blackhole: 0 -> 1

# echo net.inet.tcp.blackhole=2 >> /etc/sysctl.conf
# echo net.inet.udp.blackhole=1 >> /etc/sysctl.conf
{noformat}

One of mine is all port snapped up to date, another is further back. Both are FreeBSD 9.

Next just run one of the problem tests.

I use SolrCmdDistributorTest, because it's also what prompted the switch. This test makes
a bunch of connections very quickly. With the non NIO connector, it sometimes takes 2 seconds,
sometimes 2 minutes. With the NIO connector it's always 2 seconds in my experience. However,
on FreeBSD, you have the hangs after the test is over...
                
> Many tests on Apache Jenkins are failing with lingering threads.
> ----------------------------------------------------------------
>
>                 Key: SOLR-4006
>                 URL: https://issues.apache.org/jira/browse/SOLR-4006
>             Project: Solr
>          Issue Type: Bug
>            Reporter: Mark Miller
>
> I think I've tracked this down to being related to the black hole.
> It seems to be a recovery call to a server that is down or something - it's hanging in
the connect method even though we are using a connect timeout.
> {noformat}
> Thread[RecoveryThread,5,TGRP-SyncSliceTest]
> java.net.PlainSocketImpl.socketConnect(Native Method)
> java.net.AbstractPlainSocketImpl.doConnect(AbstractPlainSocketImpl.java:327)
> java.net.AbstractPlainSocketImpl.connectToAddress(AbstractPlainSocketImpl.java:193)
> java.net.AbstractPlainSocketImpl.connect(AbstractPlainSocketImpl.java:180)
> java.net.SocksSocketImpl.connect(SocksSocketImpl.java:384)
> java.net.Socket.connect(Socket.java:546)
> {noformat}

--
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