cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Benedict (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (CASSANDRA-9669) If sstable flushes complete out of order, on restart we can fail to replay necessary commit log records
Date Fri, 15 Jan 2016 10:42:39 GMT

    [ https://issues.apache.org/jira/browse/CASSANDRA-9669?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15101603#comment-15101603
] 

Benedict commented on CASSANDRA-9669:
-------------------------------------

The only way to make it backwards compatible is to carry on blithely if the value is not present.
 I'm not sure I'm a particular fan of this approach, because this is pretty critical to correctness,
and simply ignoring its absence is ripe for future bugs screwing it up.

> If sstable flushes complete out of order, on restart we can fail to replay necessary
commit log records
> -------------------------------------------------------------------------------------------------------
>
>                 Key: CASSANDRA-9669
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-9669
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Local Write-Read Paths
>            Reporter: Benedict
>            Assignee: Benedict
>            Priority: Critical
>              Labels: correctness
>             Fix For: 2.2.x, 3.0.x, 3.x
>
>
> While {{postFlushExecutor}} ensures it never expires CL entries out-of-order, on restart
we simply take the maximum replay position of any sstable on disk, and ignore anything prior.

> It is quite possible for there to be two flushes triggered for a given table, and for
the second to finish first by virtue of containing a much smaller quantity of live data (or
perhaps the disk is just under less pressure). If we crash before the first sstable has been
written, then on restart the data it would have represented will disappear, since we will
not replay the CL records.
> This looks to be a bug present since time immemorial, and also seems pretty serious.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Mime
View raw message