Conversation
Signed-off-by: bwplotka <[email protected]>
|
Would you like to try running a custom benchmark from this branch to make sure it works? |
Testing prometheus/test-infra#821 Signed-off-by: Bartlomiej Plotka <[email protected]>
|
I didn't plan this, no - wanted to spend my time somewhere else (low risk, low impact, can be fixed in seconds) 🙃 But given you pointed out, let's do this: prometheus/prometheus#15914 |
|
Well glad you did asked for it
(: |
Signed-off-by: bwplotka <[email protected]>
|
Removed PRW2 comms for now as it depends on metadata-wal-records feature which is problematic currently: prometheus/prometheus#15911 |
|
Tested prometheus/prometheus#15914 works now. |
ArthurSens
left a comment
There was a problem hiding this comment.
Thanks! Idea for the future: adding the new sink metrics/logs to a separate dashboard :)
|
Why did the filenames change? |
* Add PRW 2.0 traffic, improved sink debugging flow. Signed-off-by: bwplotka <[email protected]> * Removed PRW 2.0 as it depends on expensive metadata-wal-records feature. Signed-off-by: bwplotka <[email protected]> --------- Signed-off-by: bwplotka <[email protected]>
This improves PRW benchmark flow by:
Adding PRW 2.0 traffic which we should care about with the next Prometheus releases. We could drop PRW 1.0 communication given no development on that path, but since ppl still use it, let's keep it.I removed it for now as it depends onmetadata-wal-recordsfeature which is problematic currently: Deprecate and ultimately removemetadata-wal-recordsfeature enabled. prometheus#15911sinkwe will know which issue belongs to what Prometheus (e.g. prometheus from the PR is suddenly not sending help).