cshuo opened a new pull request, #18759:
URL: https://github.com/apache/hudi/pull/18759
### Describe the issue this Pull Request addresses
Flink minibatch with the global record level index was not fully consuming
the batched lookup result. Miss-heavy workloads could still fall back to
single-key index lookups during record processing, which weakens the intended
minibatch optimization and adds avoidable RLI lookup overhead.
This PR tightens the minibatch RLI path so changing records are assigned
using the prefetched batch lookup result, including confirmed misses, while
preserving immediate handling for index records and the non-upsert path.
### Summary and Changelog
- Refactors `BucketAssignFunction` to split index-record, changing-record,
and insert-record processing paths.
- Adds a write-operation-specific data record processor so non-index records
avoid repeated mode branching.
- Allows changing-record processing to consume a prefetched
`HoodieRecordGlobalLocation`, where `null` represents a minibatch-confirmed
miss.
- Updates `MinibatchBucketAssignFunction` to batch-fetch record locations
and pass the fetched results directly into bucket assignment.
- Keeps index records out of the minibatch buffer and applies them
immediately through `processIndexRecord`.
- Makes `GlobalRecordLevelIndexBackend#get(String)` explicitly unsupported
and relies on `get(List<String>)` for RLI lookup.
- Adjusts minibatch and datasource test setup to exercise global RLI
minibatch lookup with non-empty lookup results.
### Impact
Minibatch global RLI lookup now avoids redundant single-key lookups for
batch misses while preserving bucket assignment, update marking, and
index-record handling semantics.
<!-- Describe any public API or user-facing feature change or any
performance impact. -->
### Risk Level
Medium, because the change touches the Flink bucket assignment and global
RLI lookup path; mitigated by updated unit coverage and an integration test
path covering minibatch lookup behavior.
<!-- Accepted values: none, low, medium or high. Other than `none`, explain
the risk.
If medium or high, explain what verification was done to mitigate the
risks. -->
### Documentation Update
<!-- Describe any necessary documentation update if there is any new
feature, config, or user-facing change. If not, put "none".
- The config description must be updated if new configs are added or the
default value of the configs are changed.
- Any new feature or user-facing change requires updating the Hudi website.
Please follow the
[instruction](https://hudi.apache.org/contribute/developer-setup#website)
to make changes to the website. -->
### Contributor's checklist
- [ ] Read through [contributor's
guide](https://hudi.apache.org/contribute/how-to-contribute)
- [ ] Enough context is provided in the sections above
- [ ] Adequate tests were added if applicable
--
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]
