Send new label 'status' for ingester failures #4442
Conversation
danielblando
force-pushed
the
ingester-error-status-family
branch
from
4aeb956 to
cb69329
Compare
danielblando
force-pushed
the
ingester-error-status-family
branch
from
b06e8c2 to
3510532
Compare
danielblando
force-pushed
the
ingester-error-status-family
branch
from
848022a to
3be2730
Compare
|
@pracucci, I am wondering how come I don't see a "Approve and run" button in the workflow panel like https://docs.github.com/en/actions/managing-workflow-runs/approving-workflow-runs-from-public-forks suggests? |
danielblando
force-pushed
the
ingester-error-status-family
branch
from
3be2730 to
e829b47
Compare
danielblando
force-pushed
the
ingester-error-status-family
branch
from
e829b47 to
70a966e
Compare
|
@pracucci do you mind approving again? That would make the workflow to run |
|
Hmm not sure actually what happened. I think the PR just need an update to fire the workflow. It started after my previous comment. |
Sorry for the mystery here… I started the check manually after seeing your previous comment. |
|
Cool @pstibrany. Thanks |
bboreham
changed the title
pkg/distributor/distributor_test.go
Outdated
| @@ -2141,6 +2170,7 @@ type mockIngester struct { | |||
| client.IngesterClient | |||
| grpc_health_v1.HealthClient | |||
| happy bool | |||
| errFail error | |||
There was a problem hiding this comment.
Sorry, now I come to read this again I can't see why this is called errFail.
It looks like this is the error that will be returned if not happy - is that correct?
Logically a 4xx error ought to be associated with the request not the ingester.
There was a problem hiding this comment.
@bboreham Yes, you are right. I was trying to use the same pattern we already have for the test globally, but as this can be a 4xx or a 5xx it does not necessarily mean an error. I changed to 'failResp'. Let me know if you have another suggestion.
danielblando
force-pushed
the
ingester-error-status-family
branch
2 times, most recently
from
6ecc089 to
3ed8e0f
Compare
Signed-off-by: Daniel Blando <[email protected]>
Signed-off-by: Daniel Blando <[email protected]>
Signed-off-by: Daniel Blando <[email protected]>
Signed-off-by: Daniel Blando <[email protected]>
danielblando
force-pushed
the
ingester-error-status-family
branch
from
3ed8e0f to
3efe528
Compare
* Send new label statusFamily for ingestor failures Signed-off-by: Daniel Blando <[email protected]> * update changelog Signed-off-by: Daniel Blando <[email protected]> * Change changelog Signed-off-by: Daniel Blando <[email protected]> * Update label name Signed-off-by: Daniel Blando <[email protected]> Signed-off-by: Manish Kumar Gupta <[email protected]>
* Send new label statusFamily for ingestor failures Signed-off-by: Daniel Blando <[email protected]> * update changelog Signed-off-by: Daniel Blando <[email protected]> * Change changelog Signed-off-by: Daniel Blando <[email protected]> * Update label name Signed-off-by: Daniel Blando <[email protected]> Signed-off-by: Alvin Lin <[email protected]>
What this PR does:
Add new label
statusto metriccortex_distributor_ingester_append_failures_totalThis is intended to differentiate errors sent from ingester
Which issue(s) this PR fixes:
Fixes #4441
Checklist
CHANGELOG.mdupdated - the order of entries should be[CHANGE],[FEATURE],[ENHANCEMENT],[BUGFIX]