64bit snmp cellular interface issue

Hi There,

We are seeing an issue with the SNMP cellular interface counters on a Max BR1 Mini. Firmware 7.0.3
For each interface there are two different counters. 32bit ifInOctets/ifOutOctets and 64bit ifHCInOctets and ifHCOutOctets.
The 32bit counters are used primaraly with SNMPv1 and the 64bit with SNMPv2c and v3.

When polling both counters, we see major increases in the 64bit counter that we do not see in the 32bit counter, where we expect the increase to be the same.

I’ve written a little script (for testing) that polls the device every 30 seconds and displays the calculation for both OIDs.

In the output below you can see that in the first poll the OctDiff (Octets increased in the last 30 second) is the same for the normal and the HC counter.
In the second poll you can see that this is still correct for the Input, yet the 64bit output counter is completly wrong. It indicates a speed of 1.1Gbps.
Then at the third poll it’s normal again.

It should be easy to replicate if you poll a device with the OIDs below (ifIndex in this case is 4).

ifInOctets = .
ifOutOctets = .
ifHCInOctets = .
ifHCOutOctets = .

Calculated over last 30 seconds (first poll)
In Oct : 17041945
In OctDiff : 4382
In bps : 1168
In HCOct : 343614425625
In HCOctDiff : 4382
In HCbps : 1168
Out Oct : 40899286
Out OctDiff : 3574
Out bps : 953
Out HCOct : 4565591134934
Out HCOctDiff : 3574
Out HCbps : 953

Calculated over last 30 seconds (second poll)
In Oct : 17047129
In OctDiff : 5184
In bps : 1382
In HCOct : 343614430809
In HCOctDiff : 5184
In HCbps : 1382
Out Oct : 40902474
Out OctDiff : 3188
Out bps : 850
Out HCOct : 4569886105418
Out HCOctDiff : 4294970484
Out HCbps : 1145325462

Calculated over last 30 seconds (Third poll)
In Oct : 17049779
In OctDiff : 2650
In bps : 706
In HCOct : 343614433890
In HCOctDiff : 3081
In HCbps : 821
Out Oct : 40905881
Out OctDiff : 3407
Out bps : 908
Out HCOct : 4569886108825
Out HCOctDiff : 3407
Out HCbps : 908


Do you have more polling results that can share here ? Do you find any trends for the suspect invalid data ? It will always happen for the second polling gaps ?

1 Like


Engineering had tried to reproduce the reported issue and found that disable and enable may cause the issue. Would you please confirm ?

Tentatively the identified issue should fix in next firmware release (7.1.0 version)

1 Like


Thanks for the follow-up.

What do you mean by

disable and enable may cause the issue

Just to be clear, the interface was not disabled & enabled between the SNMP polls. I’ll gladly test the new firmware.
Do you still need more polling logs? I will be able to provide those later today.

Would you please open a support ticket here, for us to check on the device. We are looking on the potential items that will trigger the issue.

Please make sure RA access is enable for the device.

1 Like