[rhn-users] Total amount of system memory continually increases!

Nick Baronian kvetch at gmail.com
Mon Jun 26 14:32:41 UTC 2006


I am not the best at this type of analysis but you can try the following

Turn on Sys Requests
# echo "1" > /proc/sys/kernel/sysrq

For checking whats going on in memory allocation
# echo 'm' > /proc/sysrq-trigger

It might be helpful to dump the threads as well
# echo 't' > /proc/sysrq-trigger

This shows a lot of info so interpreting this info can be difficult.
Good Luck,
Nick Baronian




On 6/23/06, Philippe CAVAILLES <philippe.cavailles at tlse.sofreavia.fr> wrote:
> Dear ReadHead users,
> I use heartbeat software from Linux HA (http://www.linux-ha.org/) on my
> Linux RedHat ES3
> servers (Kernel version server1: 2.4.21-4.ELsmp and Kernel version
> server2: 2.4.21-40.ELsmp
> with 1Go of RAM).
>
> I face up to the following problem:
> When heartbeat is running the total amount of system memory continually
> increases
> (command: "free –mk" or "top – first part") but memory used by all the
> processes and threads running on
> the system is stable (heartbeat too! I use the command "top" column
> SIZE, RSS). The swap used is always
> equal to 0ko and when I stop heartbeat the total amount of system memory
> become stable again but doesn't decrease.
>
> And, when the full memory is used …….. The system freeze.
>
> I don't understand where this memory could be lost. Does someone have an
> idea?
> Does someone have met the same problem?
>
> I my mind the problem is not due to the use of heartbeat softwate but
> maybe to some
> systems or loaded modules calls triggered by heartbeat.
>
> How can I track the memory used by the kernel and the load modules
> (event if, I think that heartbeat run in user space)?
>
> Best regards,
> Philippe
>
> _______________________________________________
> rhn-users mailing list
> rhn-users at redhat.com
> https://www.redhat.com/mailman/listinfo/rhn-users
>




More information about the rhn-users mailing list