PERF: Skip header check in parquet reader - #22679
Conversation
The parquet reader previously required 3 reads to read the parquet footer: 1. A 4 byte read to check the header for the parquet magic bytes 2. An 8 byte read to read the footer length and footer parquet magic bytes 3. A varaible-length read for the footer metadata We don't really care about ensuring that the header is valid. For high-latency storage, it's not worth the extra read.
|
Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually. Contributors can view more details about this message here. |
|
/ok to test 0c744ac |
|
Actionable comments posted: 0 |
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yaml Review profile: CHILL Plan: Enterprise Run ID: 📒 Files selected for processing (1)
📝 WalkthroughSummary by CodeRabbit
WalkthroughThis PR refactors Parquet footer validation by removing an upfront header magic check in ChangesParquet Footer Validation Refactoring
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~20 minutes Suggested labels
Suggested reviewers
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
|
Actionable comments posted: 0 |
This reverts commit 269a7b9.
|
/merge |
Description
The parquet reader previously required three reads to read the parquet footer:
We don't really care about ensuring that the header is valid. For high-latency storage, it's not worth the extra read.
Part of #22668, which also proposes to remove the second 8-byte read. But this is a smaller change that should be less controversial.