PCM+ showing infinite errors

[Terry Kirk]

I've recently upgraded to V2.1 Update 1, and on traffic monitor; the Utilization and Frames/sec gauges are working OK, but the Broadcasts, Multicasts & Errors gauges all show negative numbers (-29496, -29496 & -589 respectively). They were working before the upgrade.
Now that I look at it, the values on the gauges are negative too.
The switch indicated on each of these gauges (a 2848) is not having any problems that I can see.

What's going on here?

Thanks

Terry

[Matt Hobbs · Accepted solution]

I would try deleting the device from PCM and then performing a manual discovery on the switch and re-adding it to traffic monitor... see if that helps.

Matt

[Terry Kirk]

Had to delete and rediscover two switches, but that fixed it.

Thanks

Terry

[Terry Kirk]

As noted above

Terry

[Steve Britt]

Terry,

Deleting and rediscovering the device, then re-adding it to traffic, is a good suggestion in either case but I was wondering how you got to this point.

So to clarify this, were you running PCM+ 2.1 and then you autoupdated it from the web? Or were you running an older version of PCM+ (pre-2.1) and you upgraded to 2.1 and also applied the autoupdate? If you upgraded what version of PCM+ were you running before? Did you do a database migration during the upgrade? What version of firmware are you running on the 2848? What type of port(s) are you monitoring (duplex and speed)?

The behavior you're seeing certainly sounds anomalous - of course you should never see negative values. I'm curious about the negative values you report though ... the multicast and broadcast values are the same (-29496) so are these the maximum values shown on the right of the gauges? Or are these the actual data values being reported? I was thinking you might mean they were the maximums only because you go on to say that the values are negative too. Could you clarify both of those comments please ...?

Whatever reported values you're seeing for the traffic metrics, are the counts consistent from minute to minute? Are you seeing this on one particular port or all ports of the 2848? Are you monitoring the device with sampling or statistics polling? You mention that this behavior recently started - what kind of values were you seeing reported before the update/upgrade?

I would suggest you open a support call for this. Support is free - it only requires time on your end, not money. That way we can work with you to get some trace files from PCM+ that will quickly clarify what PCM+ is receiving from the device, and hopefully we can reproduce the issue in the factory.

Regards,

SVB