cassandra-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jonathan Ellis (JIRA)" <>
Subject [jira] [Updated] (CASSANDRA-2419) Risk of counter over-count when recovering commit log
Date Tue, 03 May 2011 01:59:03 GMT


Jonathan Ellis updated CASSANDRA-2419:

    Attachment: 2419-v4.txt

I tried two ways of storing metadata as yaml -- first with the metadata as java beans that
were stored directly as yaml, and second half-manually serializing to a yaml Map<String,
Object> -- and both feel clunkier than just using the version field to deal with adding
things. (Especially when you need to do version checks anyway when modifying things rather
than just adding new fields.)

So, v4 is substantially the same as v3 but Descriptor version is bumped to g and we use that
instead of EOF to when reading RP. (Also, writeStatistics is renamed to writeMetadata.)

> Risk of counter over-count when recovering commit log
> -----------------------------------------------------
>                 Key: CASSANDRA-2419
>                 URL:
>             Project: Cassandra
>          Issue Type: Bug
>          Components: Core
>    Affects Versions: 0.8 beta 1
>            Reporter: Sylvain Lebresne
>            Assignee: Sylvain Lebresne
>              Labels: counters
>             Fix For: 0.8.0
>         Attachments: 0001-Record-CL-replay-infos-alongside-sstables-v2.patch, 0001-Record-and-use-sstable-replay-position.patch,
2419-v3.txt, 2419-v4.txt
>   Original Estimate: 8h
>  Remaining Estimate: 8h
> When a memtable was flush, there is a small delay before the commit log replay position
gets updated. If the node fails during this delay, all the updates of this memtable will be
replay during commit log recovery and will end-up being over-counts.

This message is automatically generated by JIRA.
For more information on JIRA, see:

View raw message