There are two values that get tracked here: Tx and Rx bandwidth utilization at the interface. These two values are collected from the following two SNMP MIB OID's:
If you plug these two OID's into your SNMP Real-Time Graph tool, click Add MIB and point it at your new Force10. Enter the SNMP community string and the OID's above and monitor these values. What are you seeing?
I get data, that isn't the issue. The issue is that on an HP switch the gagues (specifically the needle) doesn't go back to zero when it is polling the data. With the force10 the gagues go back to zero before displaying the new data.
When you see the gauge go back to zero, it usually indicates that the polling interval is faster than the device updates the counters. More than likely your force10 device is updating the ifInOctets and ifOutOctets values (the ones Jason mentioned) less frequently than every three seconds. These counters are ever-increasing, and we use the difference between one poll and the next to calculate bandwidth utilization.
Try setting the polling interval higher, say, 10 seconds or so.
That was it, it appears that 15 seconds is the sweet spot, anything below that it goes to zero. So the evice must update every 15 seconds.