Add gauge metrics to track missedEvents and cache sizes#5411
Add gauge metrics to track missedEvents and cache sizes#5411amartinezfayo merged 2 commits intospiffe:mainfrom
Conversation
342c902 to
21feef9
Compare
There was a problem hiding this comment.
This heavily uses the word "missed" which in my mind implies that we forgot something; because, one only misses items that exist (or existed).
I would prefer using the word "skipped" as there are scenarios where we should skip event ids, because the events associated with them never were created (due to hung transactions that never completed, or autoincrement values that don't increase by one).
Also, we need updates to doc/telemetry/telemetry.md.
35ccd20 to
53f5776
Compare
edwbuck
left a comment
There was a problem hiding this comment.
Thank you very much for these changes. They are a big contribution in monitoring a small value with big implications around the new database event cache.
53f5776 to
cadd672
Compare
zmt
left a comment
There was a problem hiding this comment.
LGTM. I left a few minor nits and some general pre-existing style questions.
a219c7e to
68047c5
Compare
Signed-off-by: stevend <stevend@uber.com>
68047c5 to
8988214
Compare
amartinezfayo
left a comment
There was a problem hiding this comment.
Thank you @stevend-uber!
Signed-off-by: stevend <stevend@uber.com> Co-authored-by: Agustín Martínez Fayó <amartinezfayo@gmail.com>
This PR adds metrics for the event-based cache. This will emit a gauge whenever the cache is updated for missed-events and current cache sizes.
Pull Request check list
Which issue this PR fixes
#4720