pjfanning commented on PR #1934:
URL: 
https://github.com/apache/pekko-connectors/pull/1934#issuecomment-5603249818

   Merged the test-cleanup work from #1935 into this branch, so this PR is now 
self-contained and does not need #1935 to land first.
   
   Why it was needed: CI runs `sbt "+elasticsearch/test"`, so the whole suite 
runs **twice** against the same servers — once on Scala 2.13.18, once on 3.3.8. 
The previous run's first pass was fully green; the second pass failed 14 Scala 
and 2 Java tests purely on state the first pass left behind, because `DELETE 
/_all` is refused from Elasticsearch 8 on (`action.destructive_requires_name` 
defaults to true) and the old cleanup discarded the response. The failures were 
all leftover-state signatures — `Object is missing required member 'price'` 
(Scala reading Java-written documents), `version conflict, document already 
exists (current version [2])`, `expected: <1> but was: <2>`.
   
   The merged fix lists indices via `_cat/indices` and deletes them by explicit 
name, and checks both responses so a cleanup that stops working fails loudly.
   
   One thing the automatic merge got wrong and I corrected: it left the V8 and 
V9 cleanup on the `_all` path — the two servers where it does not work — while 
V5 and V7 moved to the helper. All four now go through `deleteAllIndices`.
   
   Verified locally: compile on 2.13.18 and 3.3.8 (no warnings), 
`elasticsearch/mimaReportBinaryIssues` clean, scalafmt/javafmt clean.


-- 
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]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to