ding-young opened a new pull request, #17029: URL: https://github.com/apache/datafusion/pull/17029
## Which issue does this PR close? <!-- We generally require a GitHub issue to be filed for all bug fixes and enhancements and this helps us generate change logs for our releases. You can link an issue to this PR using the GitHub syntax. For example `Closes #123` indicates that this PR will close issue #123. --> - Closes #16909 . ## Rationale for this change In multi-level merge, we reserve estimated memory need for merging sorted spill files first, and bypass global memory pool when creating `SortPreservingMergeStream`(shortly SPM). The purpose of it is to ensure that we can finish `SPM` step without lacking memory by keeping worst case memory reservation til SPM ends. <details> - grow merge_reservation based on max batch memory per spill file https://github.com/apache/datafusion/blob/66d6995b8f626f28f811489bd2cb552b6c64a85f/datafusion/physical-plan/src/sorts/multi_level_merge.rs#L256-L268 - bypass global buffer pool (use unbounded memory pool) https://github.com/apache/datafusion/blob/66d6995b8f626f28f811489bd2cb552b6c64a85f/datafusion/physical-plan/src/sorts/multi_level_merge.rs#L326-L336 </details> Since we use `UnboundedMemoryPool` as a trick, we don't validate whether this `merge_reservation` is the actual upper limit when `SPM` step for multi level merge. Therefore, we need to validate the memory consumption in SPM does not exceed the size of merge_reservation. <!-- Why are you proposing this change? If this is already explained clearly in the issue then this section is not needed. Explaining clearly why changes are proposed helps reviewers understand your changes and offer better suggestions for fixes. --> ## What changes are included in this PR? This PR creates a separate `GreedyMemoryPool` size of `merge_reservation` instead of using `UnboundedMemoryPool` when merging spill files (and in-memory streams) on multi-level merge. <!-- There is no need to duplicate the description in the issue here but it is sometimes worth providing a summary of the individual changes in this PR. --> ## Are these changes tested? **Yes, and following tests related to spilling fail** :cry: Maybe our previous worst-case memory estimation was wrong, but don't understand why at this point. We need more investigation here. I'll put more details in comments. <!-- We typically require tests for all PRs in order to: 1. Prevent the code from being accidentally broken by subsequent changes 2. Serve as another way to document the expected behavior of the code If tests are not included in your PR, please explain why (for example, are they covered by existing tests)? --> ## Are there any user-facing changes? <!-- If there are user-facing changes then we may require documentation to be updated before approving the PR. --> <!-- If there are any breaking changes to public APIs, please add the `api change` label. --> -- 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]
