We are running Engineers Toolset with Performance Monitor 9.2. We have some Gb ports that show 1.2 Gb burst usage. Does anyone know the explanation for how we could get measurements higher than the rated port speed?
bwhiting - One reason might be that the traffic counters have cleared for the interface, and we are using rollover method 1 to deal with the lower counter value (it will assume that the counter has actually rolled-over rather than re-baselining the counter).
HTH,
Greg
Thanks. Is there a way to counteract the effect so that it will display the actual max values? In our case we had just restarted the application after doing a DB compact & repair to relaim space. Will it fix itself over time?
Sure - You can set to rollover method 2, which will re-sample the counters.
Neither method is exact, since we are fudging a number in someplace. We can either:
Stopping and starting NPM should not have anything to do with the issue though, as it is the counters on the actual device that have rolled over.
The counters for an interface (provided we are using the ifxTable, which are 64bit counters) will only naturally rollover every 0xffffffff ffffffff (18,446,744,073,709,551,615) bytes transferred, so it will take a while. However, if the counters on the device are cleared frequently, it can occur more often.