fix: TimeoutTicker returns wrong value/timeout pair when timeouts are scheduled at ~approximately the same time (backport #3092)#3107
Merged
sergio-mena merged 3 commits intov0.37.xfrom May 22, 2024
Conversation
… scheduled at ~approximately the same time (#3092) #3091 The problem is we have an edge case where we should drain the timer channel, but we "let it slide" in certain race conditions when two timeouts are scheduled near each other. This means we can have unsafe timeout behavior as demonstrated in the github issue, and likely more spots in consensus. Notice that aside from NewTimer and OnStop, all timer accesses are from the same thread. In NewTimer we can block until the timer is drained (very quickly up to goroutine scheduling). In OnStop we don't need to guarantee draining before the method ends, we can just launch something into the channel that will kill it. In the main timer goroutine, we can safely maintain this "timerActive" variable, and force drain when its active. This removes the edge case. The test I created does fail on main. --- #### PR checklist - [X] Tests written/updated - [x] Changelog entry added in `.changelog` (we use [unclog](https://github.com/informalsystems/unclog) to manage our changelog) - [x] Updated relevant documentation (`docs/` or `spec/`) and code comments - [X] Title follows the [Conventional Commits](https://www.conventionalcommits.org/en/v1.0.0/) spec (cherry picked from commit 153281a) # Conflicts: # consensus/ticker.go # consensus/ticker_test.go
Contributor
Author
|
Cherry-pick of 153281a has failed: To fix up this pull request, you can check it out locally. See documentation: https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/reviewing-changes-in-pull-requests/checking-out-pull-requests-locally |
… scheduled at ~approximately the same time (backport #3092) (#3106) The problem is we have an edge case where we should drain the timer channel, but we "let it slide" in certain race conditions when two timeouts are scheduled near each other. This means we can have unsafe timeout behavior as demonstrated in the github issue, and likely more spots in consensus. Notice that aside from NewTimer and OnStop, all timer accesses are from the same thread. In NewTimer we can block until the timer is drained (very quickly up to goroutine scheduling). In OnStop we don't need to guarantee draining before the method ends, we can just launch something into the channel that will kill it. In the main timer goroutine, we can safely maintain this "timerActive" variable, and force drain when its active. This removes the edge case. The test I created does fail on main. --- - [X] Tests written/updated - [x] Changelog entry added in `.changelog` (we use [unclog](https://github.com/informalsystems/unclog) to manage our changelog) - [x] Updated relevant documentation (`docs/` or `spec/`) and code comments - [X] Title follows the [Conventional Commits](https://www.conventionalcommits.org/en/v1.0.0/) spec <hr>This is an automatic backport of pull request #3092 done by [Mergify](https://mergify.com). --------- Co-authored-by: Dev Ojha <ValarDragon@users.noreply.github.com> Co-authored-by: Sergio Mena <sergio@informal.systems>
sergio-mena
approved these changes
May 22, 2024
PaddyMc
pushed a commit
to osmosis-labs/cometbft
that referenced
this pull request
May 23, 2024
… scheduled at ~approximately the same time (backport cometbft#3092) (cometbft#3107) cometbft#3091 The problem is we have an edge case where we should drain the timer channel, but we "let it slide" in certain race conditions when two timeouts are scheduled near each other. This means we can have unsafe timeout behavior as demonstrated in the github issue, and likely more spots in consensus. Notice that aside from NewTimer and OnStop, all timer accesses are from the same thread. In NewTimer we can block until the timer is drained (very quickly up to goroutine scheduling). In OnStop we don't need to guarantee draining before the method ends, we can just launch something into the channel that will kill it. In the main timer goroutine, we can safely maintain this "timerActive" variable, and force drain when its active. This removes the edge case. The test I created does fail on main. --- #### PR checklist - [X] Tests written/updated - [x] Changelog entry added in `.changelog` (we use [unclog](https://github.com/informalsystems/unclog) to manage our changelog) - [x] Updated relevant documentation (`docs/` or `spec/`) and code comments - [X] Title follows the [Conventional Commits](https://www.conventionalcommits.org/en/v1.0.0/) spec <hr>This is an automatic backport of pull request cometbft#3092 done by [Mergify](https://mergify.com). --------- Co-authored-by: Dev Ojha <ValarDragon@users.noreply.github.com> Co-authored-by: Sergio Mena <sergio@informal.systems> Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
Merged
7 tasks
mergify bot
added a commit
to osmosis-labs/cometbft
that referenced
this pull request
May 23, 2024
… scheduled at ~approximately the same time (backport cometbft#3092) (cometbft#3107) cometbft#3091 The problem is we have an edge case where we should drain the timer channel, but we "let it slide" in certain race conditions when two timeouts are scheduled near each other. This means we can have unsafe timeout behavior as demonstrated in the github issue, and likely more spots in consensus. Notice that aside from NewTimer and OnStop, all timer accesses are from the same thread. In NewTimer we can block until the timer is drained (very quickly up to goroutine scheduling). In OnStop we don't need to guarantee draining before the method ends, we can just launch something into the channel that will kill it. In the main timer goroutine, we can safely maintain this "timerActive" variable, and force drain when its active. This removes the edge case. The test I created does fail on main. --- #### PR checklist - [X] Tests written/updated - [x] Changelog entry added in `.changelog` (we use [unclog](https://github.com/informalsystems/unclog) to manage our changelog) - [x] Updated relevant documentation (`docs/` or `spec/`) and code comments - [X] Title follows the [Conventional Commits](https://www.conventionalcommits.org/en/v1.0.0/) spec <hr>This is an automatic backport of pull request cometbft#3092 done by [Mergify](https://mergify.com). --------- Co-authored-by: Dev Ojha <ValarDragon@users.noreply.github.com> Co-authored-by: Sergio Mena <sergio@informal.systems> Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com> (cherry picked from commit a77f195)
7 tasks
PaddyMc
pushed a commit
to osmosis-labs/cometbft
that referenced
this pull request
May 23, 2024
… scheduled at ~approximately the same time (backport cometbft#3092) (cometbft#3107) (#68) cometbft#3091 The problem is we have an edge case where we should drain the timer channel, but we "let it slide" in certain race conditions when two timeouts are scheduled near each other. This means we can have unsafe timeout behavior as demonstrated in the github issue, and likely more spots in consensus. Notice that aside from NewTimer and OnStop, all timer accesses are from the same thread. In NewTimer we can block until the timer is drained (very quickly up to goroutine scheduling). In OnStop we don't need to guarantee draining before the method ends, we can just launch something into the channel that will kill it. In the main timer goroutine, we can safely maintain this "timerActive" variable, and force drain when its active. This removes the edge case. The test I created does fail on main. --- #### PR checklist - [X] Tests written/updated - [x] Changelog entry added in `.changelog` (we use [unclog](https://github.com/informalsystems/unclog) to manage our changelog) - [x] Updated relevant documentation (`docs/` or `spec/`) and code comments - [X] Title follows the [Conventional Commits](https://www.conventionalcommits.org/en/v1.0.0/) spec <hr>This is an automatic backport of pull request cometbft#3092 done by [Mergify](https://mergify.com). --------- Co-authored-by: Dev Ojha <ValarDragon@users.noreply.github.com> Co-authored-by: Sergio Mena <sergio@informal.systems> Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com> (cherry picked from commit a77f195) Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
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.
#3091
The problem is we have an edge case where we should drain the timer channel, but we "let it slide" in certain race conditions when two timeouts are scheduled near each other. This means we can have unsafe timeout behavior as demonstrated in the github issue, and likely more spots in consensus.
Notice that aside from NewTimer and OnStop, all timer accesses are from the same thread. In NewTimer we can block until the timer is drained (very quickly up to goroutine scheduling). In OnStop we don't need to guarantee draining before the method ends, we can just launch something into the channel that will kill it.
In the main timer goroutine, we can safely maintain this "timerActive" variable, and force drain when its active. This removes the edge case.
The test I created does fail on main.
PR checklist
.changelog(we use unclog to manage our changelog)docs/orspec/) and code commentsThis is an automatic backport of pull request #3092 done by [Mergify](https://mergify.com).