Fix memory leak #51
Merged
Fix memory leak #51
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
The `Tasker::Task` that pings each websocket was not being cleaned up and garbage-collected, which meant that the WebSocket itself was also not being GCed since it (and the WebSocketPinger itself) was referenced inside the task block, leading to a leak for every single WebSocket. I had a feeling this was the case because CPU and RAM consumption were not static after a lot of WebSockets had been connected, even if they were *no longer* connected. Turns out every single WebSocket was still being pinged or raising an exception if it were closed. That exception was not causing the `Tasker::Task` to stop.
Closed
jgaskins
commented
Oct 5, 2022
src/cable/websocket_pinger.cr
Outdated
| @@ -23,10 +24,15 @@ module Cable | |||
| end | |||
|
|
|||
| def initialize(@socket : HTTP::WebSocket) | |||
| Tasker.every(Cable::WebsocketPinger.seconds.seconds) do | |||
| @task = Tasker.every(Cable::WebsocketPinger.seconds.seconds) do | |||
| Logger.debug { "Pinging websocket" } | |||
There was a problem hiding this comment.
I can remove this line if desired
There was a problem hiding this comment.
makes sense, probably would generate a log of log that brings not much value, 500 clients connected, 500 lines every second
|
I was totally thinking this same thing! On #48 there's also a RedisPinger, and I was starting with that, but I think I got tripped up by how that one was working. I wonder if it's even needed 🤔 In any case, this is amazing! Thanks for jumping in on this. |
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Outdated suggestions cannot be applied.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
Tasker::Taskthat pings each websocket was not being cleaned up and garbage-collected, which meant that the WebSocket itself was also not being GCed since it (and the WebSocketPinger itself) was referenced inside the task block, leading to a leak for every single WebSocket.I had a feeling this was the case because CPU and RAM consumption were not static after a lot of WebSockets had been connected, even if they were no longer connected. Turns out every single WebSocket was still being pinged or raising an exception if it were closed. That exception was not causing the
Tasker::Taskto stop.I discovered this by putting some debug logging inside the task block (which I forgot to remove … oops) and performed the very definitely scientific test I outlined in #50 — mashing refresh a bunch of times. With only one WebSocket open to the server (the rest had been closed), I should have seen a log message once every 3 seconds, I was seeing hundreds of them per second. With this patch, I saw them at the expected cadence.
Fixes #50