-
Notifications
You must be signed in to change notification settings - Fork 354
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Resume related Netty ByteBuf leak #779
Comments
Hi @piotr-signalfx! Thanks for reporting that. Also, you may find similar the following issue #733. In any case, we are working on improvements toward catching possible leaks in #777. However, we cannot say anything until we know that is the case that reproduces your leak. Cheers, |
Hi Oleh, We use rsocket-java 1.0.0-RC5. Cheers, |
Definitely, an example of a code that simulates what you do may simplify the investigation dramatically. In any case, we will investigate more. It might be some race cases which is not releasing all the buffers in the resuming queue |
Also, please try RC7 and see whether it has the same issue as RC5 |
Thanks. I will upgrade to RC7 and let you know if we find something more. |
Just FYI, RC-7 is in the snapshot. Briefly looking into resumability implementation, I can see a few cases which might produce leaks so I will try to work around those shortly. |
I can confirm the same issue affects RC-7 (480b501). Logs from RC-7:
|
@OlegDokuka After upgrading to RC7 we are still facing some leaks. It looks like they might be related to this error we see in the logs as the leak is always reported a few seconds later.
And the leak log:
|
@piotr-signalfx I was able to reproduce that as well locally, but unfortunately at this stage, we can nothing to do with that because of a few bugs in Reactor and in our Resumability impl as well. The good news is that the issue will be addressed in 1.1, thus, please stay tuned! |
We observe the following leak reported by Netty from time to time:
The text was updated successfully, but these errors were encountered: