How to check router interface utilization High or low

[Bijay Singh]

How to check router interface utilization High or low

[Leo Laohoo]

Post the complete output to the command "sh interface <PORT>".

[balaji.bandi]

in addition to what Leo suggested command :

more focus on :   reliability 255/255, txload 1/255, rxload 1/255 (this give you more indication what level of usage that interface)

Also other information you get more visibility if you have NMS and polling over 24hours give you peak time usage.

[Bijay Singh]

How can I check in persentage 

[Leo Laohoo]

[Joseph W. Doherty]

Do a computational, similar as I described in my prior reply, then multiply by 100.

With regard to @balaji.bandi reply, divide 1st number by 2nd number (always 255) and multiply by 100.  BTW for load values to be accurate, interface needs to be using correct bandwidth setting.

[MHM Cisco World]

Check 

show interface summary 

MHM

[liviu.gheorghe]

In addition to what @Leo Laohoo and @balaji.bandi , txload and rxload is relevant if you configure the correct interface bandwidth:

c1111#conf t
Enter configuration commands, one per line. End with CNTL/Z.
c1111(config)#interface gi0/0/1

c1111(config-if)#bandwidth 500000

This is necessary because you can connect to your ISP using a gigabit interface, which has a default bandwidth of 1000000 Kbit/sec and your contracted rate could be 500 Mbps.

Hope this helps.

[Joseph W. Doherty]

Record interface transmitted and/or received at some point of time, do again at some later point of time, subtract the former from the latter and compare to the maximum amount of bytes that could be transmitted and/or received for the same time period. Oh, and for transmitted, you may also want to count dropped bytes during the same time period added to transmitted counts.  This would provide an "offered" stat.

[Ramblin Tech]

A packet-based interface (vs a TDM interface) has only two states: 100% utilized (busy) and 0% utilized (idle). 10% utilization of an interface means that for 10% of the time during the measurement period, the interface was 100% utilized. So, was the interface utilization high or low? Intuitively, we say the utilization was low, because 90% of the time it was idle. Yet, an interface that is 90% idle can still experience queue drops because all new traffic that arrives at the interface while it is busy must be queued, which can lead to queue drops if that traffic arrives in a burst.

All that to say: do not focus exclusively on "utilization" as a performance metric. Applications do not care about "utilization", they care about loss, latency, and jitter (some apps more than others). Utilization is an interesting data point, but also be sure to measure network performance as experienced by the apps with IP-SLA/TWAMP/ThousandEyes/Accedian/etc. As as @Joseph W. Doherty alludes to: keep an eye on the packet drops as a performance metric, as drops are diagnostic of queueing and its consequent latency and latency variability (ie, jitter), not to mention the impact of the drops themselves on app performance.