Open rishabhmaurya opened 2 days ago
Hello! We have added a performance benchmark workflow that runs by adding a comment on the PR. Please refer https://github.com/opensearch-project/OpenSearch/blob/main/PERFORMANCE_BENCHMARKS.md on how to run benchmarks on pull requests.
:x: Gradle check result for c4d07359a1726c5bc80be0839a8b36265489dc77: FAILURE
Please examine the workflow log, locate, and copy-paste the failure(s) below, then iterate to green. Is the failure a flaky test unrelated to your change?
:x: Gradle check result for fe262f287a4cdef73a470cc14ea4cdd40d179ab4: FAILURE
Please examine the workflow log, locate, and copy-paste the failure(s) below, then iterate to green. Is the failure a flaky test unrelated to your change?
:x: Gradle check result for 3295caf5ac06d126e4a07507aa058c17a5f78e68: FAILURE
Please examine the workflow log, locate, and copy-paste the failure(s) below, then iterate to green. Is the failure a flaky test unrelated to your change?
:white_check_mark: Gradle check result for d053e4ac92b521fc29b7aa10fbf7d0179862554e: SUCCESS
Attention: Patch coverage is 67.39130%
with 15 lines
in your changes missing coverage. Please review.
Project coverage is 72.16%. Comparing base (
05513df
) to head (b7612f6
).
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
🚨 Try these New Features:
:white_check_mark: Gradle check result for b7612f6b748a20508c95f12bee5751be39964853: SUCCESS
On the first pass, it looks more like an SPI than a library (there is no implementation provided), where the implementations would live? Does it make sense to split it into arrow-spi
/ arrow-core
modules?
@reta you're right. Other than StreamTicket
everything else are interfaces. I like the idea of naming it arrow-spi
.
All implementations will be part of module:arrow-flight-rpc
its currently part of my feature branch.
Description
Library changes for Apache Arrow integration. Lib changes just exposes POJOs for creation of
StreamProducer
and registering them withStreamManager
.StreamProducer
in turn exposesBatchedJob
which are based on creation and filling Arrow Vectors in a batched manner handling client backpressure. So the arrow APIs exposed are kept minimal, limited to -server module will depend on
libs:arrow
to create and registerStreamProducer
for search results by populating vectors with a well defined schema for search results.Future PR will contain a module
modules:arrow-flight-rpc
, which will actually expose the Arrow Flight server, client and actual logic to create FlightStream. It will be a bulky modules in terms of all its direct and transitive dependencies.Related Issues
Resolves https://github.com/opensearch-project/OpenSearch/issues/16679
Check List
- [ ] API changes companion pull request created, if applicable.- [ ] Public documentation issue/PR created, if applicable.By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license. For more information on following Developer Certificate of Origin and signing off your commits, please check here.