Hi Team,
As per below link, when we run cbstats command it should show the bg_load
metrices value but when we check in our cluster it is not at all printing the bg_load
metrices name itself.
Could you please suggest if we are missing anything or is there any other command to print the bg_load
metrics and its value.
./cbstats xx.xx.xx.xx:11210 -u <username> -p <password> -b data1 timings
Thanks,
Debasis
I don’t see bg_load in the example in the documentation either. Maybe it only outputs statistics with non-zero values.
Please find the snippet from the documentation link I shared earlier. My concern is why the metric name does not appear when we trigger the cbstats
command in our cluster.
Thanks,
Debasis
Right. So go back to that documentation page and scroll down to the Example. There is no bg_load in their output either. Maybe it only outputs stats with non-zero values.
@mreiche Thanks, we are able to print it. Is there anyway we can only print bg_load
and bg_wait
metrices as the cbstat
is printing all cb metrices. I had tried with grep and awk to filter only these two metrices but it is not working as expected.
Thanks,
Debasis
I find jq is useful for parsing out sections of the json returned.
@mreiche The value of cbstat is not returned in JSON format. Will jq
be helpful?.
I noticed that even without any GET/SET
operations on my bucket, bg_load
and bg_wait
still print the histograms. Is the CB engine caching the previous data and printing it each time we trigger cbstat
?
If yes, can we reset the cbstats
value before starting the next round of tests on the CB cluster. This way, any data we capture will relate only to the current load..
Thanks,
Debasis
@mreiche Please ignore the request since I got the solution.
Thanks,
Debasis