nick-boss-tech commented on PR #5030: URL: https://github.com/apache/solr/pull/5030#issuecomment-6334665709
🤖 *AI text below* 🤖 *(posted on behalf of Nick Shanin)* Thanks, Eric. I read both of your commits. The extraction and the move into RemoteRequestProxy look right, the logging path keeps its behavior, and the converted cloud test passes 4 of 4 at the new head. On #5110, I agree the order does not matter: after your second commit the shared class also carries failedNodes, so it is no longer byte-identical to #5110's version, and whoever rebases second resolves that one file mechanically. I pushed one commit on top here, f8b0bbce53: /node/system now copies the proxy's failed nodes into its response, so the reporting half of your second commit lands there too, and that endpoint's cloud test asserts the field now. The Crave failure at your head traced to something unrelated to either change. Both failures were class teardown failures, and they are one event: when the client close pool in CoreContainer.shutdown() times out, the throw skips the closes that follow it, and the container's metrics contexts leak; the 16 leaked contexts were then charged to the next suite in the fork. I filed this as [SOLR-18564](https://issues.apache.org/jira/browse/SOLR-18564) and opened the fix as [#5192](https://github.com/apache/solr/pull/5192), with a test that reproduces that exact failure on current code and passes with the fix. -- 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]
