Hi:
We are using Orion NPM version 8.1.0
We have a gigabit ethernet over which a large amount of data are transferred. During this transfer, the actual interface util is observed around 80% (800mbps). This util is observed on the device and by a 3rd party snmp poller/charting program.
However, on Orion, we see only 100mbps at most during the transfer. We know this is wrong. But when there is not much traffic on the link, Orion chart on this interface appears to be accurate.
I followed the thread below and checked the snmp poller packets with ethereal.
It clearly shows it polls only if-mibs, rather than ifx-mibs.
Clearly, this error is caused by using ifTable. Orion should use ifXTable on this device/interface.
When the load is 800Mb/s, the 5 minute (our poll interval) Out_TotalByte should be around 3*10^10. This causes 32bit counters rollover multiple times.
This device is a Cisco WS-C6509. version 6.4. The ifXTable is supported by the device as a 3rd party snmp poller can poll it.
So, my question is: how to make sure the ifXTabel is polled? Also, how to make sure it is set correctly for all devices? as we have quite some interfaces that could cause ifmibs rollover.
