perf: whole-file resolution for batch UPDATE fastPatch reads - #374
Merged
Conversation
UpdateMultiple loads the whole plaintext legacy variable-length file once (<=32MB) and serves the fastPatch raw-row reads from that snapshot (one slice copy per row) instead of a pread pair per updated row. Positions that already carry a buffered in-place overwrite in this transaction are detected via the new IStorage.HasBufferedOverwriteAt and always fall back to the per-record read (which honors the write-behind buffer), so in-batch repeated updates stay correct. Measured (Release, comparative UPDATE 10K of ~100K rows, median of 3): SQL ~35K -> ~47K ops/s, Direct ~49K -> ~66K ops/s.
|
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Wat\nUpdateMultiple laadt het hele plaintext legacy variabele-lengte .dat een keer (<=32MB) en bedient de fastPatch raw-row reads vanuit die snapshot (1 slice-copy per rij) i.p.v. een pread-paar per rij. Posities met een in-batch buffered overwrite worden via de nieuwe IStorage.HasBufferedOverwriteAt gedetecteerd en vallen altijd terug op de per-record read (die de write-behind buffer eert).\n\n## Meting (Release, UPDATE 10K van ~100K rijen; median van 3)\nSQL ~35K -> ~47K ops/s; Direct ~49K -> ~66K ops/s.\n\n## Validatie\nFull suite (Debug) + Release/CI-filter op alle vier de suites: EXIT=0.