[
https://issues.apache.org/jira/browse/CASSANDRA-2816?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13054289#comment-13054289
]
Peter Schuller commented on CASSANDRA-2816:
-------------------------------------------
I've thought about this problem too, and it is really significant for some
use-cases. Again because so few writes are needed in order to trigger large
amounts of data being sent given the merklee tree granularity.
While I'm all for a fixing it by e.g. more immediate snapshotting, I would like
to raise the issue that repairs overall have pretty significant side-effects;
particularly ones that can self-magnify and cause further problems. Beyond the
obvious "it does disk I/O" and "It uses CPU", we have:
* Over-repair due to merklee tree granularity can cause jumps in CF sizes,
killing cache locality
* Combine that with concurrent repairs then repairing the "size-jumped" set of
sstables and you can magnify that effect on other nodes causing huge size
increases.
* Up to recently, mixing large and small cf:s was a significant problem if you
wanted to have different frequencies and different gc grace times, due to one
repair blocking on another. But fixes to this and the other JIRA about
concurrency, might disable the "fix" for that that was concurrent compaction -
so back to square one.
I guess overall, it seems very easy to shoot yourself in the foot with repair.
Any opinions on CASSANDRA-2699 for longer term changes to repair?
> Repair doesn't synchronize merkle tree creation properly
> --------------------------------------------------------
>
> Key: CASSANDRA-2816
> URL: https://issues.apache.org/jira/browse/CASSANDRA-2816
> Project: Cassandra
> Issue Type: Bug
> Components: Core
> Affects Versions: 0.7.0, 0.8.0
> Reporter: Sylvain Lebresne
> Assignee: Sylvain Lebresne
> Labels: repair
>
> Being a little slow, I just realized after having opened CASSANDRA-2811 and
> CASSANDRA-2815 that there is a more general problem with repair.
> When a repair is started, it will send a number of merkle tree to its
> neighbor as well as himself and assume for correction that the building of
> those trees will be started on every node roughly at the same time (if not,
> we end up comparing data snapshot at different time and will thus mistakenly
> repair a lot of useless data). This is bogus for many reasons:
> * Because validation compaction runs on the same executor that other
> compaction, the start of the validation on the different node is subject to
> other compactions. 0.8 mitigates this in a way by being multi-threaded (and
> thus there is less change to be blocked a long time by a long running
> compaction), but the compaction executor being bounded, its still a problem)
> * if you run a nodetool repair without arguments, it will repair every CFs.
> As a consequence it will generate lots of merkle tree requests and all of
> those requests will be issued at the same time. Because even in 0.8 the
> compaction executor is bounded, some of those validations will end up being
> queued behind the first ones. Even assuming that the different validation are
> submitted in the same order on each node (which isn't guaranteed either),
> there is no guarantee that on all nodes, the first validation will take the
> same time, hence desynchronizing the queued ones.
> Overall, it is important for the precision of repair that for a given CF and
> range (which is the unit at which trees are computed), we make sure that all
> node will start the validation at the same time (or, since we can't do magic,
> as close as possible).
> One (reasonably simple) proposition to fix this would be to have repair
> schedule validation compactions across nodes one by one (i.e, one CF/range at
> a time), waiting for all nodes to return their tree before submitting the
> next request. Then on each node, we should make sure that the node will start
> the validation compaction as soon as requested. For that, we probably want to
> have a specific executor for validation compaction and:
> * either we fail the whole repair whenever one node is not able to execute
> the validation compaction right away (because no thread are available right
> away).
> * we simply tell the user that if he start too many repairs in parallel, he
> may start seeing some of those repairing more data than it should.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
