[
https://issues.apache.org/jira/browse/CASSANDRA-15213?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17014704#comment-17014704
]
Benedict Elliott Smith edited comment on CASSANDRA-15213 at 1/13/20 10:59 PM:
------------------------------------------------------------------------------
Just to reiterate, though, it's still not a great test, and it might be
worthwhile improving it first. Ideally we would have tighter control on the
amount of competition we're measuring, since that's what we're interested in.
We probably also do not want to visit the same values in the same order for
every invocation, as this can lead to increased synchrony in execution lowering
overall throughput artificially. We also probably _do_ want to use the same
random sequence on each run, to improve reproducibility.
was (Author: benedict):
Just to reiterate, though, it's still not a great test, and it might be
worthwhile improving it first. Ideally we would have tighter control on the
amount of competition we're measuring, since that's what we're interested in.
We probably also do not want to visit the same values in the same order for
every execution, as this can lead to increased synchrony in execution lowering
overall throughput artificially. We also probably _do_ want to use the same
random sequence on each run, to improve reproducibility.
> DecayingEstimatedHistogramReservoir Inefficiencies
> --------------------------------------------------
>
> Key: CASSANDRA-15213
> URL: https://issues.apache.org/jira/browse/CASSANDRA-15213
> Project: Cassandra
> Issue Type: Bug
> Components: Observability/Metrics
> Reporter: Benedict Elliott Smith
> Assignee: Jordan West
> Priority: Normal
> Fix For: 4.0-beta
>
>
> * {{LongAdder}} introduced to trunk consumes 9MiB of heap without user
> schemas, and this will grow significantly under contention and user schemas
> with many tables. This is because {{LongAdder}} is a very heavy class
> designed for single contended values.
> ** This can likely be improved significantly, without significant loss of
> performance in the contended case, by simply increasing the size of our
> primitive backing array and providing multiple buckets, with each thread
> picking a bucket to increment, or simply multiple backing arrays. Probably a
> better way still to do this would be to introduce some competition detection
> to the update, much like {{LongAdder}} utilises, that increases the number of
> backing arrays under competition.
> ** To save memory this approach could partition the space into chunks that
> are likely to be updated together, so that we do not need to duplicate the
> entire array under competition.
> * Similarly, binary search is costly and a measurable cost as a share of the
> new networking work (without filtering it was > 10% of the CPU used overall).
> We can compute an approximation floor(log2 n / log2 1.2) extremely cheaply,
> to save the random memory access costs.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org
For additional commands, e-mail: commits-h...@cassandra.apache.org
