cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Branimir Lambov (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 Thu, 20 Aug 2015 11:21:48 GMT

    [ https://issues.apache.org/jira/browse/CASSANDRA-9669?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14704693#comment-14704693
] 

Branimir Lambov commented on CASSANDRA-9669:
--------------------------------------------

The code looks good now, +1.

bq. the ranges are inclusive-start, exclusive-end, the inverse of what you expected

Yes, they are the inverse of what I expected, but that's actually start-exclusive, end-inclusive,
the [shouldReplay comment|https://github.com/apache/cassandra/commit/459b96333c84837fe757d706c2a6be91b8b27f2e#diff-2b76f3efa4aaa38339ab8f4869c9b7bfR88]
is correct. I'd also add "as a mutation's replay position is assigned after it is added to
the log" to it.

Deleting the files at [the start of the test|https://github.com/apache/cassandra/commit/459b96333c84837fe757d706c2a6be91b8b27f2e#diff-b3802331f8dcba05356ad47ee54fe6dfR321]
may cause problems on Windows, but I don't know how to this safely in 2.0.


> 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: Core
>            Reporter: Benedict
>            Assignee: Benedict
>            Priority: Critical
>              Labels: correctness
>             Fix For: 3.x, 2.1.x, 2.2.x, 3.0.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