[
https://issues.apache.org/jira/browse/HBASE-10007?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13828011#comment-13828011
]
Nick Dimiduk commented on HBASE-10007:
--------------------------------------
Okay [~jmspaggi] I'll post a patch for 0.94 as well. We can address bringing
0.94's version up to feature parity with 0.96/trunk in a separate ticket.
Yes, [~apurtell], I considered that. I'm not a statistician, so I don't know if
that approach would be sufficient to keep tabs on the extremes. I'm also
interested to pursue the technique employed by Gil's
[LatencyUtils|http://latencyutils.github.io/LatencyUtils/] project. Unless
someone has immediate advice on the sampling implementation, I'd prefer to
implement this as a separate ticket.
> PerformanceEvaluation: Add sampling and latency collection to randomRead test
> -----------------------------------------------------------------------------
>
> Key: HBASE-10007
> URL: https://issues.apache.org/jira/browse/HBASE-10007
> Project: HBase
> Issue Type: Improvement
> Components: Performance, test
> Reporter: Nick Dimiduk
> Assignee: Nick Dimiduk
> Priority: Minor
> Fix For: 0.98.0, 0.96.1
>
> Attachments: HBASE-10007-0.96.00.patch, HBASE-10007-0.96.01.patch,
> HBASE-10007.00.patch
>
>
> As I mentioned over on HBASE-9940, I'd like to have randomRead operate only
> on a sample of the total dataset. It would also be useful to collect latency
> measurements from individual responses. Throughput times are aggregated
> according to the amount of user data processed and the result is reported as
> well. This is a patch I've been using for some performance tests I've run --
> maybe it'll be useful to someone else.
--
This message was sent by Atlassian JIRA
(v6.1#6144)
