[
https://issues.apache.org/jira/browse/CASSANDRA-8496?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15306552#comment-15306552
]
Branimir Lambov commented on CASSANDRA-8496:
--------------------------------------------
Option 1 is probably untenable -- we want to be able to discard the log
segments that contain those system table mutations. The other thing is that a
solution that only solves one half of the problem isn't really what we should
be after. Over that I'd most probably prefer to invest in making sure missing
the blocking flush isn't that easy.
Longer term I think we need to define a metadata mutation journal of some sort
with an association between replay positions and metadata states (which is
along the lines of your Option 2, but quite a bit more involved).
> Remove MemtablePostFlusher
> --------------------------
>
> Key: CASSANDRA-8496
> URL: https://issues.apache.org/jira/browse/CASSANDRA-8496
> Project: Cassandra
> Issue Type: Improvement
> Reporter: Benedict
> Assignee: Branimir Lambov
> Priority: Minor
>
> To improve clearing of the CL, prevent infinite growth, and ensure the prompt
> completion of tasks waiting on flush in the case of transient errors, large
> flushes or slow disks, in 2.1 we could eliminate the post flusher altogether.
> Since we now enforce that Memtables track contiguous ranges, a relatively
> small change would permit Memtables to know the exact minimum as well as the
> currently known exact maximum. The CL could easily track the total dirty
> range, knowing that it must be contiguous, by using an AtomicLong instead of
> an AtomicInteger, and tracking both the min/max seen, not just the max. The
> only slight complexity will come in for tracking the _clean_ range as this
> can now be non-contiguous, if there are 3 memtable flushes covering the same
> CL segment, and one of them completes later. To solve this we can use an
> interval tree since these operations are infrequent, so the extra overhead is
> nominal. Once the interval tree completely overlaps the dirty range, we mark
> the entire dirty range clean.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
