[Backport 2.19] Monitoring: Accurately capture predict failures (#4525)#4841
[Backport 2.19] Monitoring: Accurately capture predict failures (#4525)#4841pyek-bot wants to merge 1 commit into
Conversation
) (opensearch-project#4555) * fix: correctly capture ml failure stats on predict failures * refactor: remove redundant braces * refactor: dont track any 400 errors and add positive test case * fix: apply spotless --------- (cherry picked from commit a28d344) Signed-off-by: Pavan Yekbote <pybot@amazon.com> Co-authored-by: Pavan Yekbote <pybot@amazon.com> Signed-off-by: Pavan Yekbote <pybot@amazon.com>
56e412f to
69e8df5
Compare
PR Reviewer Guide 🔍(Review updated until commit 69e8df5)Here are some key observations to aid the review process:
|
PR Code Suggestions ✨Latest suggestions up to 69e8df5
Previous suggestionsSuggestions up to commit 69e8df5
|
|
Persistent review updated to latest commit 69e8df5 |
|
Spotless failing due to: |
Summary
shouldTrackRemoteFailure(Exception e)method that skips tracking for user/configuration issues (4xx HTTP errors,IllegalArgumentException) while still tracking infrastructure/service issues (5xx errors, other exceptions)runPredictthat previously hard-codedfalse(never tracking remote failures)Test plan
testShouldTrackRemoteFailurevalidates all scenarios