Gaps in grahps for Incomming traffic

[FreRange]

Hi,

I'm having an issue with gapping in my graphs for incomming traffic on my routers out in the field. I'm not too sure where I should start looking for the issue to correct it and I need a little help from you all.

Attached is a sample for the graphing issue that I'm having.

Windows 2003 server w/ SP1
MySQL 5
RRD 1.2
Cacti 8.6h

[BSOD2600]

100 / 1000 MB interface? Should be using snmpv2 and switch to 64bit counters.

[FreRange]

100 / 1000 MB interface? Should be using snmpv2 and switch to 64bit counters.

Nope... Cisco 2610 or a Bay ARN. This only seems to be happening on my routers, servers graph just fine. Very odd.

[BSOD2600]

It's still possible that the routers are passing enough traffic to roll the 32bit snmpv1 counters. Try using snmpv2 and the 64bit graphs.

Another possiblity is the devices are timing out with the snmp queries. Try increasing the snmp timeout for the problem routers. Could also change the logging level of cacti to medium for a while and watch when you have a problem. Then look in cacti/log file and see if cacti considers them down.