Arbitrary number based on a percentage of the link. I've never really paid any attention to those on the VLAN level unless the device is experiencing a performance issue.
255/255 is 100% utilization of a link, but most of the time even if you have full utilization you'll see a reported value of something like 248+/255. I cannot remember if I have ever seen 255/255.
It's just a percentage, fwiw.
I guess I am just trying to figure out how to effectively monitor my links in Orion. My Vlan 1 interface is always 100% but Orion thinks that is based on 1Gbps...
The reliability of an interface is always reported as a fraction of 255 (255/255 is 100% reliability), calculated as an exponential average over 5 minutes. So you want to see the reliability at 255/255. If this drops below this value, then you have problems on the link. Typically, any layer 1 or layer 2 issues will bring this number down say 228/255.
The load on an interface is also a fraction of 255 (255/255 is completely saturated), and is also calculated as an exponential average over 5 minutes. You do NOT want to see the TX load or RX load at 255/255 or 100%, as this indicates a saturated link. The configured bandwidth on the interface and the actual bandwidth utilized affect this number. For example, you have a 100Mbps link and it is 50% utilized, you should see something like 128/255.
Message was edited by: Brian
Ok and back to the what is bandwidth of a vlan? It all comes back to that. This isnt like a WAN link I need to monitor, I need to monitor traffic utilized on a vlan.
ip address 10.151.1.1 255.255.0.0
ip helper-address 10.100.51.10
ip helper-address 10.10.10.14
ip nat inside
ip pim sparse-mode
ip flow ingress
ip flow egress
So what is the bandwidth command on this vlan interface used for?
I could be wrong, but wouldn't the "traffic used on a VLAN" be the aggregate of traffic load across all ports assigned to the VLAN? The Vlan1 interface is just a virtual interface, not an actual load-bearing interface.
Why are you tracking the utilization of a VLAN?? That is my question. Track the bottlenecks - which are the physical links. Then track the CPU utilization to monitor overall box health.
Did you issue this command under the interface? If so, why? Below is the output from one of my distribution switches, notice there is no bandwidth configured under the interface. However, in the output of the "sh interface" command, you see a bandwidth of 100,000 kbps or 100 Mbps.
ip address 192.168.11.1 255.255.255.0
DSW-1#sh int vlan 11
Vlan11 is up, line protocol is up
Hardware is CPU Interface, address is 0000.0c95.4c48 (bia 0000.0c95.4c48)
Internet address is 192.168.11.1/24
MTU 1500 bytes, BW 100000 Kbit, DLY 1000000 usec,
reliability 255/255, txload 1/255, rxload 1/255
I would say this is just a default bandwidth assigned to the interface just as what you get when you create a loopback interface. Notice the "delay" on the interface as well. Again, these are just defaults added by the IOS for such interfaces as loopbacks, tunnels, SVI, DVI, etc. Remember what the bandwidth command is used for? QoS and routing protocols.
I agree with Travis here. Since you mention you have many users coming in on VLAN1 and your servers are connected on this VLAN as well, try setting up Orion to monitor the "servers" ports. These are the ports that are going to get overloaded as many users try to access them. As the traffic load on a single server interface gets above 60% (continuous utilization) then upgrade to an Etherchannel (if 100Mb links) or 1Gb links (configure Etherchannel once utilization goes above 60%) then upgrade to 10Gb links for the servers.
Make sure CEF is enabled if you ar running any sorta recent ios. 'ip cef' must be globally enabled for netflow to be working. just my 2 cents worth of b/w here!