[jira] [Commented] (CASSANDRA-9504) Odd performance numbers comparing 2.0.15 and 2.1+

["Philip Thompson (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/CASSANDRA-9504?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14567889#comment-14567889 ] 

Philip Thompson commented on CASSANDRA-9504:
--------------------------------------------

It doesn't appear there was a notice in NEWS.txt for 2.1 for the change in batch commitlog behavior. Is this something worth considering adding (even if a little late)? Since it does appear that someone upgrading from 2.0 to 2.1 would need to make changes to their yaml to get similar performance behavior/avoid a serious performance regression.

> Odd performance numbers comparing 2.0.15 and 2.1+
> -------------------------------------------------
>
>                 Key: CASSANDRA-9504
>                 URL: https://issues.apache.org/jira/browse/CASSANDRA-9504
>             Project: Cassandra
>          Issue Type: Improvement
>            Reporter: Philip Thompson
>             Fix For: 3.x, 2.2.x
>
>         Attachments: jstack.tar.gz
>
>
> I've been doing some basic performance testing of batch commitlog on 2.0.15, 2.1-head, and 2.2.-head. I have been seeing very odd results where 2.0.15 performance is incredibly high, suspiciously so, when compared to 2.1 or 2.2.
> Example graph:
> http://cstar.datastax.com/graph?stats=2ea11748-0562-11e5-9045-42010af0688f&metric=op_rate&operation=1_write&smoothing=1&show_aggregates=true&xmin=0&xmax=1570.36&ymin=0&ymax=127509.8
> I'm using the following yaml options:
> {code}
> commitlog_sync: batch
> commitlog_sync_batch_window_in_ms: 50
> commitlog_sync_period_in_ms: null
> concurrent_writes: 64
> {code}
> I am able to reproduce this on two separate clusters, physical hardware, with SSDs. I have not been able to reproduce this on gce, or a physical cluster I have access to with HDD. I can reproduce this consistently, with different stress options, and with higher data load on the nodes. I can also reproduce this myself on these machines, without using the C* perf tool to set up the cluster and perform the operations.
> When running 2.0.15 batch commitlog vs periodic, I do see a difference as expected:
> http://cstar.datastax.com/graph?stats=4ae77aac-056f-11e5-8344-42010af0688f&metric=op_rate&operation=1_write&smoothing=1&show_aggregates=true&xmin=0&xmax=68.09&ymin=0&ymax=166966.8
> I am using the newest stress code on trunk. What should I be looking for to explain the anomalous performance here?



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)