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

FS#833 - pppoe memory problem #7994

Closed
openwrt-bot opened this issue Jun 9, 2017 · 5 comments
Closed

FS#833 - pppoe memory problem #7994

openwrt-bot opened this issue Jun 9, 2017 · 5 comments
Labels

Comments

@openwrt-bot
Copy link

edy-c:

I have observed a problem with pppoe connection. PPP runs well, but if the connection is interrupted by synchronization etc. then the free memory is less.

If I pull the plug and wait, then the system rapidly loses free memory and then crashes by approximately 25 MB of free RAM.

echo 3> /proc/sys/vm/drop_caches does not help in this case.

Tested on Linksys WRT1200AC, kernel 4.4.69 and 4.4.70
LEDE 17.01

Is there a way to investigate the problem more closely?

Many Thanks

Edy

@openwrt-bot
Copy link
Author

lucize:

I can see the same problem with WRT3200ACM, if pppoe connection is setup, not having service, the router will run out of memory, like ~2MB on each pppoe session, mine will not crash but will become unusable with very slow response

SNAPSHOT, r4855-45d0b88

@openwrt-bot
Copy link
Author

lucize:

Tried on x86 and sun7i arm board with 4 pppoe sessions ~30 min and no memory leak (just 2MB of increased memory usage) so maybe the leak is related to other network components that get restarted (switch drivers) ?

@openwrt-bot
Copy link
Author

edy-c:

There was no memory leaks on the Linksys WRT1200AC router with OpenWRT Chaos Calmer 15.05.1.

It looks as if the source code of ppp was not changed, so the problem is to be found elsewhere.

@openwrt-bot
Copy link
Author

olenchenko:

I have the same issue when pppoe tries to reconnect using Linksys WRT1900ACS, LEDE Reboot 17.01.4 r3560-79f57e422d, Kernel Version 4.4.92.
echo 3> /proc/sys/vm/drop_caches does not help.

Yesterday my ISP had some major issues and my 1900ACS become unusable after ~45mins - I could not connect to it (web and SSH) and had to reboot it with power switch.

Is there a way to debug this problem?
This issue looks to me like a critical one.
Any advice?

@openwrt-bot
Copy link
Author

edy-c:

For the moment you can try to add the following to /etc/sysctl.conf

vm.overcommit_memory = 2
vm.overcommit_ratio = 95

and type sysctl -p to apply or reboot.

With this settings, the kernel will prevent applications to allocate more memory than available.

In our case this will not work and the router will reboot before it becomes unusable after all memory is occupied.
If you have a separate syslog server you can see what the kernel try to do to stop that applications.

Maybe someone of LEDE developers reads this here and are willing to help to solve this problem.

Thanks

Edy

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

No branches or pull requests

1 participant