mattcasters commented on PR #11660: URL: https://github.com/apache/nifi/pull/11660#issuecomment-5604245129
Thanks Pierre. The PR is based on current `main` (rebased onto `NIFI-16319` / `b6a04999e8` at open time). That already includes both [NIFI-16226](https://issues.apache.org/jira/browse/NIFI-16226) (`778dd665b6`) and [NIFI-16286](https://issues.apache.org/jira/browse/NIFI-16286) (`14fa8a2fb0`). `main` has since gained only `NIFI-16320`, which is unrelated. [NIFI-16326](https://issues.apache.org/jira/browse/NIFI-16326) was filed against **2.11.0**, which does not have NIFI-16226 (that landed for 2.12.0). I agree NIFI-16226 is in the same area: update analysis was reclassifying inherited Parameters as local. After that change, `setEffectiveParameterUpdates()` sets `inherited` from the source context, and `getParameters()` already skipped `inherited=true`. For the common REST path (empty parameter list, or inherited Parameters with the flag set), that should already prevent the first inheritance change from being saved as a local Parameter on `main`. What this PR still covers on current `main` is the apply path when the effective value is present on the update DTO **without** `inherited=true`, but with a source Parameter Context other than the one being updated. In that case `main` still persists the value locally, and later reorders / add-remove keep the stale effective value. That is the DAO test `testSubsequentInheritanceReorderDoesNotPersistEffectiveValueWhenInheritedFlagMissing`. I have not re-run the original UI steps against a vanilla `main` build. If you already see the Jira reproduction fixed on `main` by NIFI-16226, I am happy to close this, or keep the extra DAO guard plus the Jira-step tests as regression coverage. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
