aglinxinyuan commented on PR #6182: URL: https://github.com/apache/texera/pull/6182#issuecomment-4888262662
On the benchmark flags — these are runner noise, not a regression. This PR only bumps the `botocore`/`boto3`/`aiobotocore` family (+ the LICENSE-binary sync); none of those are on the Arrow Flight data plane — they back the S3/Iceberg storage path, which the throughput/latency micro-benchmark doesn't exercise at all. The flagged deltas bear that out: the "🔴 worse" entries are +1–15% latency on a benchmark whose 7-day average swings ±100%+ (bs=10 p95 7d avg is +116.8%), and the throughput deltas are −1.5% to −9.7%, all in the ⚪ <±5% or small-noise band. The longest, least-noisy case (bs=1000) is essentially flat: −1.5% throughput, +1.3% p50. Nothing here is attributable to a dependency-only change off the hot path. _🤖 Addressed by [Claude Code](https://claude.com/claude-code)_ -- 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]
