I currently collect ~100k traces a minute. Is ther...
# support
o
I currently collect ~100k traces a minute. Is there some way to aggregate this data after some time to not keep all traces? Or is the proper way to use another sampler when collecting traces? I want to be able to track the traffic in an accurate way.
Also, what would be the proper way to truncate the db? I tried setting the retention period to 1h to clear old data but I cant see disk space shrinking even though it says "Your last call to change retention period to 1 hr was successful.".
s
Is there some way to aggregate this data after some time to not keep all traces?
We already do this and calculate all RED metrics that power the
Services
tab.
Or is the proper way to use another sampler when collecting traces?
By collecting traces, I imagine you are referring to HEAD, which can be very inflexible sometimes because the important trace could not have been sampled in the beginning.
what would be the proper way to truncate the db
Truncate the individual tables.
I tried setting the retention period to 1h to clear old data but I cant see disk space shrinking even though it says "Your last call to change retention period to 1 hr was successful."
IIRC it is a day precision because mutations are costly and doing them often only leads to unnecessary complications.