Export client request latencies as histograms #90
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This adds support for histogram-style metrics instead of using summaries. It means we can sum on a cluster level and present the user's experienced latency instead of looking at it on a per-node level. The current version limits the range of histogram buckets between 0.1ms and 60s, to avoid exporting huge amounts of buckets that are likely empty. Further patches could limit this further, for example by going for a 1.44x increment instead of the 1.2x increment, or by specifying the ranges in the configuration. Even with the limits in place, this exports 76 metrics to 3 metric families per histogram. The original summaries-based code only exports 8 metrics (in 3 families), though in theory those are no longer needed and could be disabled with a flag.
johndelcastillo
force-pushed
the
master
branch
2 times, most recently
from
3b88272 to
84cb7cd
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This adds support for histogram-style metrics instead of using
summaries. It means we can sum on a cluster level and present the user's
experienced latency instead of looking at it on a per-node level.
The current version limits the range of histogram buckets between 0.1ms
and 60s, to avoid exporting huge amounts of buckets that are likely
empty. Further patches could limit this further, for example by going
for a 1.44x increment instead of the 1.2x increment, or by specifying
the ranges in the configuration.
Even with the limits in place, this exports 76 metrics to 3 metric
families per histogram. The original summaries-based code only exports 8
metrics (in 3 families), though in theory those are no longer needed and
could be disabled with a flag.