Nagios Support Forum

i will revisit MaxBytes - and would be very pleased if I was incorrect - sounds like I may be - thank you

on the default port speed - I believe this to be exactly that - default and yes it can be modified and no, it doesn't seem to matter once the SNMP query picks up a value from the device which cannot be changed
as to having not seen it before - this is a known bug on Cisco 10Gig ports and has been since 2012 - I have seen similar issues in the more distant past - the further back, the more there were

Thanks both though - I will work through this tomorrow and will update

joehamm wrote:i will revisit MaxBytes - and would be very pleased if I was incorrect - sounds like I may be - thank you

on the default port speed - I believe this to be exactly that - default and yes it can be modified and no, it doesn't seem to matter once the SNMP query picks up a value from the device which cannot be changed
as to having not seen it before - this is a known bug on Cisco 10Gig ports and has been since 2012 - I have seen similar issues in the more distant past - the further back, the more there were

Thanks both though - I will work through this tomorrow and will update

Makes sense then, as I have not done it on any 10G ports yet.

joehamm wrote:Thanks both though - I will work through this tomorrow and will update

Great. I hope it works for you. It may be worthwhile to open a feature request at http://tracker.nagios.com for the ability to alter the default speeds on the second page of the switch/router wizard.
Best of luck!

Guys,

Thanks again for the insights and assistance.
Having deleted all device related files and entries in the mrtg.cfg, rediscovering and changing the MaxBytes..... which did get propagated to mrtg.cfg

for some reason did not work

I investigated further only to find that the issue is more widespread (at least with all my Cisco's that have 10G interfaces)
Instead of reading octets and presenting the information based on that delta -Nagios still seems to want to report these interfaces based on the 4.29G reported from the MIB.

Based on show interface stats on the Cisco - I can easily match the Nagios value * 4.29
Might do some tcpdumps to see what else Nagios queries when collecting octets
Might do some other research

I may look at creating my own utilization graph (service) where I can specify what is being presented based on octets only
but that will take some time

If you have any other suggestions / ideas - I am game for a simpler solution
Thanks
Joe

It seems like a bug with the hardware as this should report the link speed, not max capable for the port?

as I mentioned - this is a known bug on Cisco 10G ports
but I will come up with some kind of work around
Thanks again

No problem.
You could always look at altering the MIB . . .

thanks but ifspeed is a read only mib
Object, ifSpeed. OID, 1.3.6.1.2.1.2.2.1.5. Type, Gauge32. Permission, read-only

Is there an OID that reports the right data (for link speed) ? We could replace the OID in the MIB with one that is proper.

NOTE: This may not be a good idea because it might break any other devices that do not have this issue.

as it may only be important during the discovery process - this might work
I knew about the other ifspeed oid for Cisco but had no idea where I could change in Nagios
if you have that info - I will attempt
could make for a win