Improve range request skipping with configurable buffered skip - #103
Draft
KochTobi wants to merge 3 commits into
Draft
Improve range request skipping with configurable buffered skip#103KochTobi wants to merge 3 commits into
KochTobi wants to merge 3 commits into
Conversation
Wrap the stream in a BufferedInputStream only during skipToStart so that seeking to a large offset (e.g. 90% of a multi-GB file) discards data in skipBufferSize chunks instead of byte-by-byte. The buffer is discarded immediately after the skip, so the producer-consumer streaming path and its backpressure behavior are unchanged. Make the skip buffer size configurable via server.download.skip-buffer-size (default 8192 bytes).
Log skip progress at TRACE level in skipBufferSize steps so logs confirm the skip advances in buffer-sized chunks rather than byte-by-byte. The logging is opt-in via the SKIP_TRACE_LEVEL env var (default info), leaving the root log level untouched.
The previous trace logged pseudo-progress at buffer boundaries, which is misleading when InputStream.skip() is honored: a single skip() call can discard the whole range in one jump, so no intermediate steps appear. Now each honored skip() call is logged with its byte count and offset, and the final line reports how many byte-by-byte reads fell back, making the slow path obvious.
|
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
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Wrap the stream in a BufferedInputStream only during skipToStart so that seeking to a large offset (e.g. 90% of a multi-GB file) discards data in skipBufferSize chunks instead of byte-by-byte. The buffer is discarded immediately after the skip, so the producer-consumer streaming path and its backpressure behavior are unchanged.
Make the skip buffer size configurable via
server.download.skip-buffer-size (default 8192 bytes).