fix(sockets): correctly handle pending writes to prevent double write #214
Conversation
This file contains 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
TKussel
approved these changes
Nov 8, 2024
Comment on lines
+467
to
+468
| let data: Arc<Vec<_>> = (0..13337).map(|_| { | ||
| let mut chunk = vec![0; OsRng.gen_range(1..9999)]; |
There was a problem hiding this comment.
Can you motivate these magic numbers?
There was a problem hiding this comment.
They are only in the tests, so not critical
There was a problem hiding this comment.
9999 because its larger than 8192 which is the default buffer size of tokio::io::copy_bidirectional and 13337 because its a cool number and approximately the number of writes that SEL did in my test case.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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.
The previous implementation called
sendon the sink directly without using theSinkWriteradapter to convert it to anAsyncWrite. This meant polling the returned future ourself which seemed to work fine for the majority of cases as it either returnedPoll::Readywhen we were able to send the item to the underlying io orPoll::Pendingwhen the io was busy matching the expected behavior of theAsyncWritetrait. However there is one other case where the future returned bysendmight returnPoll::Pendingwhich is when it was able to send the data to the io but flushing the underlying io returnedPoll::Pending. This is an unfornuate property of the send function which I did not realize when initially implementing it. In pratice this leads to a lot of flushes and in my tests caused the bug that occasionally a packet was sent twice as it has been written to the io but not flushed meansing the future returnedPoll::Pendingleading to the buffer being sent a second time.TLDR:
The new implementation uses the
SinkWriterwhich both does not flush on every write and ensures data is written properly.