Ping forensics reveal network performance

Few people realise the wealth of information this superstar protocol can provide

Ask just about any IT professional and they will confirm that Ping has been a part of their troubleshooting arsenal for as long as they care to remember. There is little dispute that ping is a powerful tool to test connectivity between devices, but few people realise the wealth of information this superstar protocol can provide.

What is Ping?

Ping is a networking utility provided with most operating systems including Windows and Linux. It’s commonly used for checking whether other computers are alive but it also provides timing feedback that contains a wealth of information about your network performance.

Who invented Ping

Very interesting story, nobody really knows and the Internet is ripe with speculation but the most likely culprit is an Australian man called….oh wait. I’m thinking of Bitcoin. Ping was invented by Mike Muuss in December 1983.

What can it do for you?

It can reveal problems with your service provider network or your WAN. The great part is that you don’t need to know about contention ratios, SNR, committed rates and other WAN technologies. If there is something wrong network-wise Ping will flag it for you.

What should you look for

Essentially you will be interested in high or inconsistent delay. As a rule of thumb delays are approximately;

5-20ms RTT (Round Trip Time) to your service provider
~40ms RTT to the other side of the country
~200ms RTT to the other side of the world

What do inconsistent results reveal and why

There are four primary contributors to delay;

Distance delay – Mr Einstein kindly confirmed that this delay is not negotiable. This is how long it takes for data to reach its destination at the speed of light.

Serialization delay – It probably has some fancy fandangle name these days, but us old timers know it as serialization delay. This is the time it takes for your data to jump from your network device into the cable. This type of delay can only be influenced by increasing link speed – for example a 100Mbs link will send the data on its way 10 times faster than a 10Mbs link.

Queuing delay – This is the one to watch out for. Queuing occurs when there is congestion. Just like a traffic jam on the highway, congestion occurs when the cars at the back are travelling faster than the cars at the front. This signifies a bottleneck and this causes your delay to fluctuate dramatically. If you are seeing consistent fluctuating delays it may be time to send a please explain to your WAN or service provider. To understand what is ‘normal’ on your network test it at a time you know is not busy such as late at night. Then test again at peak times and compare.

Compute delay – This is a fancy way of saying that the computer you are pinging is really, really busy. This can occur on and servers that require a lot of compute power and can sometimes be misinterpreted as network problems.

What the big deal with delay?

TCP is probably the most conscientious protocol ever written. If it detects what it deems to be slow links or congestion this good Samaritan is designed to slow your transmission rate down.

Milliseconds in terms of round trip time can manifest as seconds from a user experience point of view. Add some packet loss into the mix and that’s the recipe for lots of service desk calls from unhappy customers.

Google the term ‘TCP Throughout Calculator’ and you will find RTT (Round Trip Timer) is one of just three metrics used to calculate how quickly your computer is able to send.

You may also be interested in

The role of Information Technology in Business
How Will Netflix Change the Way DTH Business Works?