fdleak somewhere

[radare]

looks like generic_fuzzer.py is leaking fds somehow

[Tue Nov  4 23:28:23 2014 25176:139983205721856] Exception: [Errno 24] Too many open files: '/proc/31950/task'
Error: [Errno 24] Too many open files: '/proc/31950/task'

[joxeankoret]

Is an unknown problem of VTrace :/ I have to check it to determine where is the bug.

[radare]

I had suffered other hangs in the balakalaaaavalaakad daemon. In my experience all queue services use to hang soon or later, but those bugs are not in your code and they are hard to reproduce.

On 05 Nov 2014, at 10:03, Joxean notifications@github.com wrote:

Is an unknown problem of VTrace :/ I have to check it to determine where is the bug.

— Reply to this email directly or view it on GitHub.

[joxeankoret]

My guess is that your VM or hardware machine has very little memory. Am I wrong? I doubt there are real problems with Beanstalkd. For example: it's used to handle all the traffic that VirusTotal receives.

[v-p-b]

I was able to resolve this with ulimit -n that indicates that the number of fd's is big but finite.

[joxeankoret]

Thanks v-p-b! I will add a page in the wiki with typical problems and not this one.

[radare]

it’s limited to 1024 , which is probably a low value and it catches the fdleak bug sooner than in other systems probably.

This is not really a solution to the problem.

On 08 Nov 2014, at 17:41, buherator notifications@github.com wrote:

I was able to resolve this with ulimit -n that indicates that the number of fd's is big but finite.

— Reply to this email directly or view it on GitHub https://github.com/joxeankoret/nightmare/issues/16#issuecomment-62264271.

[joxeankoret]

The solution to this problem is to either fix VTrace (go figure...) or drop it and use another debugging interface. I prefer to go for the 2nd and, meanwhile, offer a workaround.