[
https://issues.apache.org/jira/browse/CASSANDRA-11853?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15297454#comment-15297454
]
Nitsan Wakart commented on CASSANDRA-11853:
-------------------------------------------
So, with 100K ops/sec you hit a 150ms STW pause. The pause makes 15K ops queue
up, 30 per thread. When the pause ends, all threads go as fast as they can to
catch up. At this point they are going as hard as throughput. If max throughput
for the system is 200K per second, the buildup alone will take 75ms to clear.
Assuming some extra fallout from this and some contention as 500 threads
scramble I can see how the end result will be an observed 300ms response time
for one of them.
> Improve Cassandra-Stress latency measurement
> --------------------------------------------
>
> Key: CASSANDRA-11853
> URL: https://issues.apache.org/jira/browse/CASSANDRA-11853
> Project: Cassandra
> Issue Type: Improvement
> Components: Tools
> Reporter: Nitsan Wakart
> Assignee: Nitsan Wakart
> Fix For: 3.x
>
>
> Currently CS reports latency using a sampling latency container and reporting
> service time (as opposed to response time from intended schedule) leading to
> coordinated omission.
> Fixed here:
> https://github.com/nitsanw/cassandra/tree/co-correction
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
