How to see the bandwidth usage of router interfaces in real time?

[XavierMarichal]

Hello everyone, I want to see the bandwidth usage of router interfaces Cisco 7200 in real time is there a command to see this or maybe a calculation method? I try with SNMP and Netflow but I need the value of bandwidth usage in real time.

Thank you veyr much for your help! 

[Dennis Mink]

the sh int xyz command would give you the thru put.  network management platforms typically poll like every 1 or 5 minutes and graph accordingly. alternatively you can run a on demand snmp poller, because snmp can definitely give you the correct thruput OID. what networks management platform are you using/?

[balaji.bandi]

as suggested other post you need monitoring system in place to monitor the intercface using SNMP / Netflow depends your requirement, you want to in real time, you can set the time to poll how soon you want to results, standard to 60secm you can reduce this one, but you see other issue, monitoring system keep polling to busy device. if this is high processing traffic device you see other issues here.

what is the reason you looking this in real time ? 

[Joe Clarke]

You can set load-interval to 30 seconds on the interfaces (default is five minutes), and then you can see the 30 second bandwidth values in the "show interface" output:

  30 second input rate 0 bits/sec, 0 packets/sec
  30 second output rate 0 bits/sec, 0 packets/sec

It's not real-time, but there is nothing that will give you true real-time.  You can also use the txload and rxload meters to determine the bandwidth usage percentage for that same 30 second interval.  You can then use a small tclsh script to calculate that:

set res [exec "show int Fa0/0 | inc load"]

regexp "txload (\d+).*rxload(\d+)" $res -> txload rxload

set tx [expr ${txload}.0 / 255.0]

set rx [expr ${rxload}.0 / 255.0]

puts "TX bandwidth percentage is [expr ${tx} * 100.0]%"

puts "RX bandwidth percentage is [expr ${rx} * 100.0]%"

[xAdventix]

My answer would have been the same as Joe's, as he's correct. One thing I utilize is LibreNMS which can actually poll real-time traffic as quickly as every 0.25 seconds! 
This is a link for reference to one of the guys over there showing how it's done as evidence:

https://community.librenms.org/t/how-to-monitoring-the-interface-in-real-time-librenms-centos7/1084

[Joe Clarke]

The problem is many of our devices don't update their interface counters except every 5 or 10 seconds.

[xAdventix]

Oh, this is a good point, fair enough then, I didn't realize this, but at least you can get better than the 30seconds then? Or does cisco only poll every 30seconds making any third party product not be able to do it quicker? :)

[Joe Clarke]

You can poll more frequently with SNMP, understanding that that has an impact on the network element resources.  With the adoption of streaming, model-driven telemetry, this is improving.  Devices are beginning to support on-change telemetry updates for counters so you can have the device push that data to you.  The ASR9K is an example of such a device.  MDT is going to greatly enhance network monitoring operations as the streams can be multiplexed to different consumers thus reducing overall resource impact to the network elements themselves.

[Abrar Malik]

conf t

go to WAN interface and type the command  below .

TCL_Pri_ILL_30mbps#sho ip int brief
Interface IP-Address OK? Method Status Protocol
GigabitEthernet0/0/0 14.140.90.26 YES NVRAM up up
GigabitEthernet0/0/1 14.141.68.73 YES NVRAM up up
GigabitEthernet0/0/2 unassigned YES NVRAM administratively down down
GigabitEthernet0 unassigned YES NVRAM administratively down down
TCL_Pri_ILL_30mbps#

TCL_Pri_ILL_30mbps#sho int gig0/0/0
GigabitEthernet0/0/0 is up, line protocol is up
Internet address is 14.140.90.26/30
MTU 1500 bytes, BW 100000 Kbit/sec, DLY 100 usec,
reliability 255/255, txload 5/255, rxload 22/255
Encapsulation ARPA, loopback not set
Keepalive not supported
Full Duplex, 100Mbps, link type is force-up, media type is RJ45
output flow-control is on, input flow-control is on
ARP type: ARPA, ARP Timeout 04:00:00
Last input 00:09:10, output 00:09:10, output hang never
Last clearing of "show interface" counters never
Input queue: 0/375/0/0 (size/max/drops/flushes); Total output drops: 127
Queueing strategy: fifo
Output queue: 0/40 (size/max)
5 minute input rate 8750000 bits/sec, 1640 packets/sec        (here u can find actually bandwidth rate input and output  
5 minute output rate 1969000 bits/sec, 1002 packets/sec        you can calculate to Maps)
183470673 packets input, 133912230521 bytes, 0 no buffer
Received 24 broadcasts (0 IP multicasts)
0 runts, 0 giants, 0 throttles
0 input errors, 0 CRC, 0 frame, 0 overrun, 0 ignored
0 watchdog, 4 multicast, 0 pause input
142189686 packets output, 58603777720 bytes, 0 underruns
0 output errors, 0 collisions, 2 interface resets
0 unknown protocol drops
0 babbles, 0 late collision, 0 deferred
5 lost carrier, 0 no carrier, 0 pause output
0 output buffer failures, 0 output buffers swapped out
TCL_Pri_ILL_30mbps#

[Joseph W. Doherty]

Firstly, you may need to define what you mean by "the value of bandwidth usage in real time" because it's always either zero or 100%, in real, real time.

If you looking for average bandwidth usage (consumption) with some very small time period, best bet might be occasional SNMP polling of the RMON Ethernet history group (if available and if time period acceptable).