[
https://issues.apache.org/jira/browse/CASSANDRA-5150?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13645627#comment-13645627
]
Pawel Mirski commented on CASSANDRA-5150:
-----------------------------------------
Tested on Windows, it seems that performance drop is minimal (about 1%).
> sstable2json doesn't check SIGPIPE
> ----------------------------------
>
> Key: CASSANDRA-5150
> URL: https://issues.apache.org/jira/browse/CASSANDRA-5150
> Project: Cassandra
> Issue Type: Bug
> Components: Tools
> Affects Versions: 2.0
> Reporter: Will Oberman
> Priority: Minor
> Labels: lhf
> Attachments: trunk-5150.txt
>
>
> I believe this explains the issue better than I can:
> http://stackoverflow.com/questions/11695500/how-do-i-get-java-to-exit-when-piped-to-head.
> Basically, I expected that if I did: "sstable2json SSTABLE | other-process",
> and other-process had issues and/or died then the sstable2json process would
> die. It doesn't.
> My workaround is using mkfifo FILE, and having sstable2json write to FILE,
> other-process read from FILE, and a 3rd overall process make sure the other
> two processes are working. But, it would be _much_ simplier if sstable2json
> failed on SIGPIPE.
> I looks like the fix is to periodically check System.out.checkError() in the
> Java.
--
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
