hello-stephen opened a new pull request, #65936:
URL: https://github.com/apache/doris/pull/65936

   ### What problem does this PR solve?
   
   Related PR: #65921
   
   The native Parquet decoder and reader microbenchmark matrix is available in 
the repository, but the community `performance` admission pipeline only builds 
the normal FE/BE artifacts and never builds or runs `benchmark_test`.
   
   ### What is changed?
   
   - Build the Release benchmark target after the existing performance FE/BE 
build, using a separate output directory so deployment artifacts remain 
unchanged.
   - Run the complete Parquet matrix as an executability smoke gate: 152 
decoder cases and 137 reader cases.
   - Fail when case registration counts drift, a benchmark process fails, the 
JSON is invalid, or any result reports `error_occurred`.
   - Publish the benchmark build log, registered case list, and raw 
decoder/reader JSON through TeamCity artifacts, including partial artifacts 
when the benchmark stage fails.
   - Preserve the existing TeamCity-side `liblance_c.a` injection point used by 
the performance compile step.
   
   The run intentionally uses a 1 ms minimum time. These measurements are 
diagnostic only and must not be used to claim a performance improvement or 
regression. A statistical performance gate needs same-agent baseline/PR 
comparison and stable variance data; this PR first establishes buildability, 
full-matrix execution, artifacts, and runtime cost in the formal pipeline.
   
   ### Check List (For Author)
   
   - [x] `bash -n regression-test/pipeline/performance/prepare.sh 
regression-test/pipeline/performance/compile.sh 
regression-test/pipeline/performance/run-parquet-microbenchmark.sh`
   - [x] Stubbed end-to-end runner validation for 152 decoder and 137 reader 
results
   - [x] Simulated the current TeamCity `liblance_c.a` script injection
   - [x] `git diff --check`
   - [ ] Community TeamCity `performance` validation
   


-- 
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]

Reply via email to