wegod commented on issue #5366:
URL: https://github.com/apache/rocketmq/issues/5366#issuecomment-1310057219
> > Will you add metric gauge about consumeOKTPS and consumeFailedTPS in
ConsumeStatus.java? These data on consumer client are very helpful to find if
there are some error cause by dirty data or virtual machine issues.
> > And will you and similar consumeOKTPS and consumeFailedTPS on producer
client? Such like producerOKTPS or producerFailedTPS. Although this is not
important as consumer TPS, it's better to help business to find who is the bad
guy that send large garbage.
> > suf
>
> Hisrograms can track the number of observations , that showing up as a
time series with a _count suffix is inherently a counter. So, to calculate
consumeOKTPS we can use the following experession:
`rate(rocketmq_process_time_count{invocation_status="success"}[1m])` to
calculate producerOKTPS we can use the following experession:
`rate(rocketmq_send_cost_time_count{invocation_status="success"}[1m])`
Thanks for your reply.
I only see rocketmq_send_cost_time metrics, will you add
rocketmq_send_msg_count in Producer Metrics?
And I see topic and client_id params in Producer Metrics. Does that means
Metrics can show topic send msg count by each producer machine?
For example, there are one producer group and four producer clients those
send msg count are 1,2,3,4. Will I see 1,2,3,4 in Metrics or only a total num
10?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
