[
https://issues.apache.org/jira/browse/CASSANDRA-264?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12725246#action_12725246
]
Jonathan Ellis commented on CASSANDRA-264:
------------------------------------------
02 reads
add test catching buggy update of header on flush; refactor so there is only
one version of code doing those writes (the correct one)
the bug was in the writeCommitLogHeader called by discard; it didn't write the
length first so readLong() on replay would get basically garbage
> log replay bugs
> ---------------
>
> Key: CASSANDRA-264
> URL: https://issues.apache.org/jira/browse/CASSANDRA-264
> Project: Cassandra
> Issue Type: Bug
> Reporter: Jonathan Ellis
> Assignee: Jonathan Ellis
> Attachments: 0001-CASSANDRA-264-don-t-rm-when-replay-fails.txt,
> 0002-add-test-catching-buggy-update-of-header-on-flush-ref.txt
>
>
> - OOM on log replay
> - log is deleted even when replay fails
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
