zeroshade opened a new pull request, #1275:
URL: https://github.com/apache/arrow-go/pull/1275

   ### Rationale for this change
   
   The macOS test jobs run on 7 GB ARM64 runners and intermittently kill 
`parquet/file` while running under the race detector. Two large-value 
regression tests each write more than 1 GB and retained all encoded output in 
memory. Package-level parallelism adds further pressure.
   
   Example failure: 
https://github.com/apache/arrow-go/actions/runs/33590782213/job/100362942846?pr=1241
   
   ### What changes are included in this PR?
   
   - Replace the retained `bytes.Buffer` output in the two large-value tests 
with a counting writer.
   - Run Parquet packages serially (`-p=1`) on Darwin to bound package-level 
memory pressure.
   - Correct both macOS job names from AMD64 to ARM64.
   
   ### Are these changes tested?
   
   - `go test -race -tags assert ./parquet/file -count=1`
   - `go test -race -tags assert,noasm ./parquet/file -count=1`
   - `bash -n ci/scripts/test.sh`
   - `git diff --check`
   
   Both formerly high-memory tests also passed independently under `-race -tags 
assert,noasm`.
   
   ### Are there any user-facing changes?
   
   No.
   


-- 
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]

Reply via email to