chore(feedback): remove user-feedback-ingest rollout flag#78097
chore(feedback): remove user-feedback-ingest rollout flag#78097
Conversation
Codecov ReportAttention: Patch coverage is ✅ All tests successful. No failed tests found. Additional details and impacted files@@ Coverage Diff @@
## master #78097 +/- ##
==========================================
+ Coverage 78.12% 78.18% +0.06%
==========================================
Files 7089 7039 -50
Lines 312504 311906 -598
Branches 51052 51012 -40
==========================================
- Hits 244132 243854 -278
+ Misses 62016 56294 -5722
- Partials 6356 11758 +5402 |
src/sentry/relay/config/__init__.py
Outdated
| "organizations:standalone-span-ingestion", | ||
| "organizations:transaction-name-mark-scrubbed-as-sanitized", | ||
| "organizations:transaction-name-normalize", | ||
| "organizations:user-feedback-ingest", |
There was a problem hiding this comment.
Is the Relay PR for this opened yet? We need to make sure that we don't accidentally drop user feedback in external Relays (i.e. internal Relays will have to hard-code this flag when propagating project config downstream).
There was a problem hiding this comment.
We need to make sure that we don't accidentally drop user feedback in external Relays (i.e. internal Relays will have to hard-code this flag when propagating project config downstream).
Sorry I don't think I understand - what conditions would this happen under?
There was a problem hiding this comment.
@aliu39 customers that run their own Relay instances do not have the updated Relay code yet, so they will still check the feature flag and drop data if it's missing.
I pushed a commit to your Relay PR to prevent that, hope that clarifies what I mean!
Ref getsentry/sentry#78097 --------- Co-authored-by: Joris Bayer <joris.bayer@sentry.io> Co-authored-by: David Herberth <david.herberth@sentry.io>
Removes this rollout flag since user feedback is GA'd. Applies the options denylist directly (it was previously checked in the flag handler).
Checks the denylist in the shim function. I think this is a better place to check it - previously we were missing a spot in post-process
PR making this change safe for internal and external (self hosted) relays: getsentry/relay#4076