Hi all,
In our Datadog Taveloka - Prod
organization, we have some custom metrics with very high number of series && are sent by our building blocks, for example:
api.(req|res|res.exc).count
> 37k metrics
rpc.(client|server).(count|exc.count|ltcy.p95)
> 236k metrics
ScheduledExecutor.QueuedCount
+ Thread.(WAITING_ALL|RUNNABLE|BLOCKED)
> 260k metrics
The api.*
and rpc.*
are used by most of the teams (or all of them) and so far there doesn't seem to be any tag that can be removed to reduce the number of series.
The ScheduledExecutor.QueuedCount
and Thread.*
metrics are also used by many teams but we are not sure whether we need to monitor all the threadpools or not.
Here is some details for the threadpool metrics: https://docs.google.com/spreadsheets/d/1sG__1VT9DHuSl0UUFgsX7vJMhLXCIcWQ_WfTVr5xg5c/edit#gid=0
Please fill the column E, F, G, H, and I to help us analyze our usages and remove unused tags (tag: threadname
).
Thank you!:bow: