feat(processor): Migrate process_replay#5580
Conversation
process_replayprocess_replay
| let datascrubbing_config = ctx | ||
| .project_info | ||
| .config | ||
| .datascrubbing_settings | ||
| .pii_config() | ||
| .map_err(|e| Error::PiiConfigError(e.clone()))? | ||
| .as_ref(); |
There was a problem hiding this comment.
Should be able to pull this out of the retain.
Possibly even the RecordingScrubber, need to check if it keeps state.
There was a problem hiding this comment.
The RecordingScrubber has some internal state sadly, but I hope that this is a none-issue once the logging confirms the hypothesis that we only get one 'replay' per envelope.
| envelope.add_item( | ||
| Item(payload=PayloadRef(bytes=b"{}\n[]"), type="replay_recording") | ||
| ) |
There was a problem hiding this comment.
Why did you add this? To get more coverage out of the integration tests?
There was a problem hiding this comment.
This was added since the new logic enforces that the recording and event should be send in the same envelope (like the devdocs already say should be the case A "replay_event" Item should always be sent with a "replay_recording" Item in the same envelope.).
The tests previously just send the event and hence failed on the new logic.
| |replays| &mut replays.replays, | ||
| |replay, _| { | ||
| let event = replay.get_event().value().ok_or(Error::NoEventContent)?; | ||
| replay::validate(event).map_err(|e| Error::InvalidPayload(e.to_string())) |
There was a problem hiding this comment.
nit: Looks like we could move replay::validate to this module, it does not normalize anything.
…ntry/relay into tobias-wilfert/feat/process_replay
There was a problem hiding this comment.
Cursor Bugbot has reviewed your changes and found 1 potential issue.
Bugbot Autofix is OFF. To automatically fix reported issues with Cloud Agents, enable Autofix in the Cursor dashboard.
Used for partially rolling out this: getsentry/relay#5580
Used for partially rolling out this: getsentry/relay#5580
Migrates the
process_replaylogic to our new processor architecture.To not take any unnecessary risk this will be a partial rollout to test on selected orgs first before deploying to everyone, since we don't have enough traffic on our 'staging' environment to use that as testing.
Added a list of follow ups to the ticket.
Related: INGEST-515