Given: each event with timestamp is already in an index.
I am looking for a way to aggregate (percentile) the frequency of events.
For example, between the hours of 8am and 5pm, what is the load in that time for n days? (1 event every x seconds 99% of the time)
I am thinking the following:
- query the events between 8am and 5pm for each day in the range
- calculate the frequency of the events (I am hoping there is some help on this)
- aggregate the frequency for all of the n days (a bucket)
- then get the percentile - and render in kibana (how to render percentiles in kibana is most helpful)
Any insights tips or help before I start a bruit force approach would be helpful.