keep trying to connect to draining QUIC servers #14863
Closed
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
Do not give up connect on servers that are in draining state. This might indicate the QUIC server restarting and the UDP packet routing still hitting the instance shutting down. Instead keep on connecting until the overall TIMEOUT fires.
Since QUIC filters no longer set `reconnect_at`, remove it. Let test_03 tests run in CI, should work better now.
bagder
approved these changes
Sep 12, 2024
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Outdated suggestions cannot be applied.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.
We once established a connect retry logic in our IP Eyeballing that made a second attempt when the filter returned a WEIRD_SERVER_REPLY. QUIC filters use that when the server is responding with a "draining" connect. This happens when the server shuts down or gracefully restarts. The problem for QUIC is that UDP packets can not be routed atomically to a new process like for TCP. Such "draining" responses seem hard to avoid.
The previous approach of just retrying once leads to timing problems. When curl is faster than the server, the second attempt will also fail and abort the connect. With this PR, the happy eyeballing will continue retrying the available addresses that report draining - until success or the overall connect timeout. This is far more reliable.
Remove the old
reconnect_atvariables in the QUIC filters that are no longer used.Enable test_03 for restarting servers in CI. Hopefully those work there reliable now.