We are looking for a way to generate an alert of network devices whose memory usage is slowly increasing.
Reason for this is that we occasionally have outages caused by memory shortage, where the only resolution seems to be to reboot the device, as it is impossible to logon to the device, even if we have out of band management.
In NPM we can notice that memory has slowly crawls up to close to 100%. On some devices it may take a week or two and on some devices it can take a year. We already have an alert on high memory usage, but we want to be alerted when memory usage slowly increases.
I believe a report with a memory usage chart over the last three months would help me catch most of the offending devices by just skimming through the report once a week but I haven't been able to figure out how to create a report for all our Cisco devices with one chart per device.
Premise is to look at utilization for the last hour/day then compare it to the average of the last week / month. If there is a delta greater than "x" - throw an alert
This request is based on a question by on september 23th 2014.
If this could be incorporated in a future version of NPM, I think it would be a great benefit for many people !
