Allowing failed splits in root search. #5440
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
fulmicoton
force-pushed
the
issue/5411-lenient-on-split-failure
branch
2 times, most recently
from
f518830 to
0dfec8f
Compare
fulmicoton
force-pushed
the
issue/5411-lenient-on-split-failure
branch
6 times, most recently
from
47abca3 to
a2401d8
Compare
This PR adds a query string parameter (`allow_partial_search_results`) in the elasticsearch api controlling encounterring an error on a subset of the splits, should fail a search query or if a response (reflecting the partial set of successful splits) should be returned. It changes the behavior of the root search gRPC call, that will not return an error in presence of partial split errors, but instead returns the list of splits that failed. This PR also changes the default behavior of the elasticsearch rest API: following elasticsearch behavior on failing shards, quickwit will allow split errors. The Quickwit API default behavior, on the other hand, is unchanged. Closes #5411
fulmicoton
force-pushed
the
issue/5411-lenient-on-split-failure
branch
from
a2401d8 to
8754e83
Compare
trinity-1686a
requested changes
Sep 25, 2024
fulmicoton
force-pushed
the
issue/5411-lenient-on-split-failure
branch
2 times, most recently
from
e178af0 to
6a6ee22
Compare
Co-authored-by: trinity-1686a <[email protected]>
fulmicoton
force-pushed
the
issue/5411-lenient-on-split-failure
branch
from
6a6ee22 to
ea024cd
Compare
trinity-1686a
approved these changes
Sep 26, 2024
Comment on lines
861
to
864
| // Ideally, we would have wanted to reuse the setting from the initial search request. | ||
| // However, passing that parameter is cumbersome and not necessary: | ||
| // if we have a valid `scroll_id` it means that the search request was successful. | ||
| // We can therefore allow failed splits, it won't make any difference. |
There was a problem hiding this comment.
i agree it's cumbersome, and i'm fine with it being left as TODO.
I'm not convinced by the argument that if we have a scroll id, search was successful. After a few pages, we need to do more search, which might hit a missing split (or fail for whatever reason), so an initial success only guarantees we're okay for so much calls.
There was a problem hiding this comment.
:( yes you are correct.
fulmicoton
force-pushed
the
issue/5411-lenient-on-split-failure
branch
from
0e87fe3 to
d922852
Compare
The logic is now aligned with what is done for queries: we check the schema at the root level. If the aggregation does not make sense for the current index docmapping, the query is rejected.
fulmicoton
force-pushed
the
issue/5411-lenient-on-split-failure
branch
from
d922852 to
8a166b7
Compare
trinity-1686a
approved these changes
Sep 27, 2024
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
(I also tested the option by removing splits on a running quickwit instance)