[
https://issues.apache.org/jira/browse/CASSANDRA-11853?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15298648#comment-15298648
]
Nitsan Wakart commented on CASSANDRA-11853:
-------------------------------------------
How would you like this to be reported?
Currently we have:
INFO 14:32:16 Latency mean : 0.8 ms [insert: 1.1 ms, simple1:
0.8 ms]
INFO 14:32:16 Latency median : 0.4 ms [insert: 0.9 ms, simple1:
0.4 ms]
INFO 14:32:16 Latency 95th percentile : 1.6 ms [insert: 2.2 ms, simple1:
1.5 ms]
INFO 14:32:16 Latency 99th percentile : 8.9 ms [insert: 6.9 ms, simple1:
9.1 ms]
INFO 14:32:16 Latency 99.9th percentile : 39.2 ms [insert: 34.2 ms, simple1:
40.0 ms]
INFO 14:32:16 Latency max : 56.8 ms [insert: 53.6 ms, simple1:
56.8 ms]
I would not recommend doing the extra reporting in place, as it will be quite
hard to have 3 measurements for total+each op in one line.
We could go for my original suggestion which was:
INFO 14:32:16 rTime mean : 0.8 ms [insert: 1.1 ms, simple1:
0.8 ms]
INFO 14:32:16 rTime median : 0.4 ms [insert: 0.9 ms, simple1:
0.4 ms]
INFO 14:32:16 rTime 95th percentile : 1.6 ms [insert: 2.2 ms, simple1:
1.5 ms]
INFO 14:32:16 rTime 99th percentile : 8.9 ms [insert: 6.9 ms, simple1:
9.1 ms]
INFO 14:32:16 rTime 99.9th percentile : 39.2 ms [insert: 34.2 ms, simple1:
40.0 ms]
INFO 14:32:16 rTime max : 56.8 ms [insert: 53.6 ms, simple1:
56.8 ms]
INFO 14:32:16 sTime mean : 0.8 ms [insert: 1.1 ms, simple1:
0.8 ms]
INFO 14:32:16 sTime median : 0.4 ms [insert: 0.9 ms, simple1:
0.4 ms]
INFO 14:32:16 sTime 95th percentile : 1.6 ms [insert: 2.2 ms, simple1:
1.5 ms]
INFO 14:32:16 sTime 99th percentile : 8.9 ms [insert: 6.9 ms, simple1:
9.1 ms]
INFO 14:32:16 sTime 99.9th percentile : 39.2 ms [insert: 34.2 ms, simple1:
40.0 ms]
INFO 14:32:16 sTime max : 56.8 ms [insert: 53.6 ms, simple1:
56.8 ms]
INFO 14:32:16 wTime mean : 0.8 ms [insert: 1.1 ms, simple1:
0.8 ms]
INFO 14:32:16 wTime median : 0.4 ms [insert: 0.9 ms, simple1:
0.4 ms]
INFO 14:32:16 wTime 95th percentile : 1.6 ms [insert: 2.2 ms, simple1:
1.5 ms]
INFO 14:32:16 wTime 99th percentile : 8.9 ms [insert: 6.9 ms, simple1:
9.1 ms]
INFO 14:32:16 wTime 99.9th percentile : 39.2 ms [insert: 34.2 ms, simple1:
40.0 ms]
INFO 14:32:16 wTime max : 56.8 ms [insert: 53.6 ms, simple1:
56.8 ms]
But I expect this will break existing tools down stream parsers.
Your call.
> 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)
