Thursday, April 10, 2025
2 changes · saas-18.2
Resolved issues and error corrections
This change makes the Discuss call system record only more serious peer-to-peer connection issues by default. It prevents expected, already handled call connection situations from filling the logging database, helping administrators focus on actionable errors.
Original PR description
Before this commit, the default log level of `peer_to_peer` was set to `WARN` which includes: - recovery candidates: recoveries can be triggered for various non critical reasons as recovery is initiated eagerly but will only realize if necessary (in which case it is logged at `ERROR` level). - reception of bus events for peers that aren't registered: can happen in race conditions that are already handled. - offer collision rollback: offer collisions can happen during recovery and fallbacks, and are already handled. This commit raises the default log level from `WARN` to `ERROR` which will prevent the aforementioned situations to be logged by default (outside of the `debug` mode) and fill the logging database.
Users now see the correct notification when trying to reschedule task dependencies involving a closed task in the Gantt view. This prevents confusing or missing error messages and helps project teams understand why the scheduling action cannot be completed.
Original PR description
Steps to reproduce: - In gantt view add two tasks and make them dependent. - Closed the source task. - Click on forward schedule(right button) on connecter. Issue: - Invaild error/ No error message. - Issue works vice verca(target closed record to be scheduled backward) also Reason: - Before we used to raise error using client actions. - But all the other errors have been changed to be using Notification Service this might have been missed. Fix: - Changed the format so it works as expected. task-4715001