[jira] [Commented] (SOLR-3685) Solr Cloud sometimes skipped peersync attempt and replicated instead due to tlog flags not being cleared when no updates were buffered during a previous replication.

2012-09-19 Thread Markus Jelsma (JIRA)
[ https://issues.apache.org/jira/browse/SOLR-3685?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13458679#comment-13458679 ] Markus Jelsma commented on SOLR-3685: - {quote}So what we're seeing here is the mmapped

[jira] [Commented] (SOLR-3685) Solr Cloud sometimes skipped peersync attempt and replicated instead due to tlog flags not being cleared when no updates were buffered during a previous replication.

2012-09-18 Thread Robert Muir (JIRA)
[ https://issues.apache.org/jira/browse/SOLR-3685?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13457914#comment-13457914 ] Robert Muir commented on SOLR-3685: --- Whats happening with this issue: is it still one? sh

[jira] [Commented] (SOLR-3685) Solr Cloud sometimes skipped peersync attempt and replicated instead due to tlog flags not being cleared when no updates were buffered during a previous replication.

2012-08-22 Thread Uwe Schindler (JIRA)
[ https://issues.apache.org/jira/browse/SOLR-3685?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13439666#comment-13439666 ] Uwe Schindler commented on SOLR-3685: - Hi, another thing to take into account: Java all

[jira] [Commented] (SOLR-3685) Solr Cloud sometimes skipped peersync attempt and replicated instead due to tlog flags not being cleared when no updates were buffered during a previous replication.

2012-08-20 Thread Markus Jelsma (JIRA)
[ https://issues.apache.org/jira/browse/SOLR-3685?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13437790#comment-13437790 ] Markus Jelsma commented on SOLR-3685: - To my surprise the RES for all nodes except the

[jira] [Commented] (SOLR-3685) Solr Cloud sometimes skipped peersync attempt and replicated instead due to tlog flags not being cleared when no updates were buffered during a previous replication.

2012-08-16 Thread Uwe Schindler (JIRA)
[ https://issues.apache.org/jira/browse/SOLR-3685?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13436243#comment-13436243 ] Uwe Schindler commented on SOLR-3685: - Hi, I also don't think MMap is the reason for th

[jira] [Commented] (SOLR-3685) Solr Cloud sometimes skipped peersync attempt and replicated instead due to tlog flags not being cleared when no updates were buffered during a previous replication.

2012-08-16 Thread Markus Jelsma (JIRA)
[ https://issues.apache.org/jira/browse/SOLR-3685?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13436214#comment-13436214 ] Markus Jelsma commented on SOLR-3685: - We didn't think mmap could be the cause but neve

[jira] [Commented] (SOLR-3685) Solr Cloud sometimes skipped peersync attempt and replicated instead due to tlog flags not being cleared when no updates were buffered during a previous replication.

2012-08-16 Thread Yonik Seeley (JIRA)
[ https://issues.apache.org/jira/browse/SOLR-3685?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13436209#comment-13436209 ] Yonik Seeley commented on SOLR-3685: May also want to try specifying NIOFSDirectoryFact

[jira] [Commented] (SOLR-3685) Solr Cloud sometimes skipped peersync attempt and replicated instead due to tlog flags not being cleared when no updates were buffered during a previous replication.

2012-08-16 Thread Markus Jelsma (JIRA)
[ https://issues.apache.org/jira/browse/SOLR-3685?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13436194#comment-13436194 ] Markus Jelsma commented on SOLR-3685: - One node also got rsyslogd killed but the other

[jira] [Commented] (SOLR-3685) Solr Cloud sometimes skipped peersync attempt and replicated instead due to tlog flags not being cleared when no updates were buffered during a previous replication.

2012-08-16 Thread Mark Miller (JIRA)
[ https://issues.apache.org/jira/browse/SOLR-3685?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13436187#comment-13436187 ] Mark Miller commented on SOLR-3685: --- Are there any crash dump files? I don't think I've s

[jira] [Commented] (SOLR-3685) Solr Cloud sometimes skipped peersync attempt and replicated instead due to tlog flags not being cleared when no updates were buffered during a previous replication.

2012-08-16 Thread Markus Jelsma (JIRA)
[ https://issues.apache.org/jira/browse/SOLR-3685?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13436182#comment-13436182 ] Markus Jelsma commented on SOLR-3685: - Finally! Two nodes failed again and got killed b