Right now my 6513 memory used is at 86 percent and slowly climbing...about 1 percent per hour. What can I look at to see what the problem maybe. Any direction would be great. Normally it runs in the 20 - 25 percent range.
I am now running at 90%. Is there a way to find out what is causing this? Also can it be resolved with out taking the switch down?
The "RPC pm-mp" is holding most of the memory, so I am assuming this is where the issue is. I would like to see a "show mem alloc total" and "show version" from the switch. A reboot of the switch (after the captures please) will temp. work around the issue as the memory will be re-carved. It will likely however start to leak again. Please provide me the information and I will look for a defect.
Thanks for your imput. I did open a TAC case as well and here is what we found. The process that was holding the memory is a process that is the RPC communication between the 2 CPU's. What we did to try and free up the memory was to force a failover between the sup's using, REDUNDANCY FORCE_SWITCHOVER, command.....that did not free up the memory. We then physically pulled the active sup which cause a fail over and then after a couple minutes reseated it and it still did not clear the memory so we pulled the other sup and reseated and that did the trick and freed up the memory. During all this we only dropped a couple pings so it did not disrupt the network traffic...which was good. They are now looking to see if there is some type of bug causing this memory leak. After we gathered some logs we saw this all started about feb. 15th or 16th and was losing about 20mb per day....it was a steady trend upwards.
Thank you very much for all your help.