cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Branimir Lambov (JIRA)" <>
Subject [jira] [Commented] (CASSANDRA-9669) If sstable flushes complete out of order, on restart we can fail to replay necessary commit log records
Date Tue, 15 Sep 2015 12:41:47 GMT


Branimir Lambov commented on CASSANDRA-9669:

doesn't look right. This guarantees that it does not contain data _before_ the given position,
and thus {{CFS.forceFlush(ReplayPosition)}} does not appear to use it correctly.

is suspect in the context of this ticket, "ka" reader says compatible with "kb" data but read
will fail. Later readers (e.g. "la") will also fail to read these tables but say they are
compatible. Could this cause trouble?

The new format needs to be added to {{LegacySSTableTest}} and {{test/data/legacy-sstables}}
and included in 2.2 and 3.0 versions of the patch.


This must satisfy {{approximateCommitLogLowerBound <= commitLogLowerBound}} (after discarding
predecessor), mustn't it? Could you add a comment to say so?

no longer loops through index CFSes, can you add a comment to state why this is the right
thing to do?

duplicate log?

[Descriptor.Version version text|]:
2._1_.7; j versions are compatible according to {{isCompatible}}, shouldn't comment be retained?

No need to declare {{commitLogUpperBound}} here. Declare with assignment below to ease reading.

Could use some text describing what's being printed.

> If sstable flushes complete out of order, on restart we can fail to replay necessary
commit log records
> -------------------------------------------------------------------------------------------------------
>                 Key: CASSANDRA-9669
>                 URL:
>             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

View raw message