Skip to content
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

Medusa gets killed due to excessive memory usage #579

Open
YoussefAbtk opened this issue Feb 11, 2025 · 12 comments
Open

Medusa gets killed due to excessive memory usage #579

YoussefAbtk opened this issue Feb 11, 2025 · 12 comments
Labels
bug Something isn't working high-priority

Comments

@YoussefAbtk
Copy link

Hi. I am using version 1.1.0 but the process is abruptly killed by my system after few seconds:

zsh: killed     medusa fuzz

Checking the logs reveals that Medusa consumes an extremely high amount of memory (~90GB), causing macOS to terminate it due to low swap space.

kernel:  kernel: process medusa[2601] caught waking the CPU 45001 times over ~175 seconds, averaging 256 wakes / second and violating a limit of 45000 wakes over 300 seconds
 kernel: low swap: killing largest compressed process with pid 2601 (medusa) and size 93366 MB

I tried to change the number of workers or even the workerResetLimit in my config file but it didn't help me at all.

@anishnaik anishnaik added bug Something isn't working high-priority labels Feb 11, 2025
@anishnaik
Copy link
Collaborator

Hi @YoussefAbtk thank you for reporting this. I actually stumbled upon this yesterday as well. I am investigating on my end where things are going wrong. Would you be comfortable sharing the source code that you used to trigger this?

@YoussefAbtk
Copy link
Author

Hi @anishnaik, Yes, of course, I’m okay with sharing the source code.

@anishnaik
Copy link
Collaborator

@YoussefAbtk thank you, really appreciate it. How long does it take for the kernel to kill the process?

@YoussefAbtk
Copy link
Author

You welcome @anishnaik, I would say 12 seconds after running the transaction in the corpus.

@anishnaik
Copy link
Collaborator

Okay, thank you so much @YoussefAbtk. Will investigate ASAP.

@YoussefAbtk
Copy link
Author

No problem, happy to help 🫡

@anishnaik
Copy link
Collaborator

@YoussefAbtk how big is the corpus btw? in terms of MB?

@YoussefAbtk
Copy link
Author

@anishnaik It's 79.9 MB for the call_sequences and 225 KB for the test_results

@anishnaik
Copy link
Collaborator

@YoussefAbtk can you do me a favor and try running Medusa v1.0.0 with the same corpus? I want to see if the OOM bug still exists in that release. If it doesn't then it would really narrow down the search space for where the bug might be.

@YoussefAbtk
Copy link
Author

@anishnaik yes of course, I’ll give it a try and get back to you.

@YoussefAbtk
Copy link
Author

@anishnaik unfortunately it's the same with this version :

zsh: killed     medusa fuzz
kernel: low swap: killing largest compressed process with pid 90714 (medusa) and size 90368 MB

@anishnaik
Copy link
Collaborator

Okay, that's good to know. Thanks @YoussefAbtk.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
bug Something isn't working high-priority
Projects
None yet
Development

No branches or pull requests

2 participants