Uh oh!
There was an error while loading. Please reload this page.
[release/6.0] [HTTP] Scavenge fix - #62008
Merged
Merged
Conversation
…longer than the timer interval gets triggered
… caller (scavenge timer callback)
ghost
commented
Nov 24, 2021
Tagging subscribers to this area: @dotnet/ncl |
geoffkizer
approved these changes
Nov 29, 2021
karelz
approved these changes
Nov 30, 2021
This was referenced Dec 16, 2021
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 freeto subscribe to this conversation on GitHub.
Already have an account?
Sign in.
5 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Backport of #61530 to release/6.0
/cc @ManickaP
Fixes#61505
Fixes#61506
Customer Impact
The issue was reported by a customer. In case the pool scavenging job would take longer than the interval in which the timer is triggered we would just start another one in parallel. The client app is a crawler visiting millions of different URLs and that's why they have so many different pools in the manager and why the scavenging job might take a while to finish.
Having multiple scavenging jobs run in parallel is unwanted (#61505). So with this change we restart the job only after the current run has finished. We also kick off connection disposal asynchronously to not to block the scavenging on actual
SslStream/NetworkStreamclosing (#61506).Testing
I've provided private binaries to the customer which confirmed (via manually collected dump after 50 mins run) that the fix helped.
Risk
Low to Medium.
It's unfeasible to have test coverage for crawler-like (millions of different URIs) scenarios that triggered the issues so we don't have our own test coverage, only validation from the customer.