Traceroute is a commonly used tool for diagnosing network routing issues, measuring delays across a network. When diagnosing various performance or network-related issues, ClearFront support will frequently ask you to run a traceroute to your hostname to help discover the source of the problem. This guide explains how to run a traceroute in just a few simple steps.

Windows

1. Press Start+R
2. Enter "cmd" and click Ok to open the command prompt
3. Now type out the command "tracert <hostname>" where <hostname> is your ClearFront hostname, for example myzone.cf-cdn.xyz
4. Press Enter on your keyboard and wait for the output to finish

Linux

1. Open a terminal window
2. Run the command "traceroute <hostname>" where <hostname> is your ClearFront hostname, for example myzone.cf-cdn.xyz

If traceroute is not installed on your system, simply install it with your system package manager.

Mac OS

1. Open the Terminal
2. Run the command "traceroute <hostname>" where the hostname is your ClearFront hostname, for example: myzone.cf-cdn.xyz

How to interpret the output and what to watch out for

A typical traceroute may look something like this:

Each line represents a ‘hop’, or part of the routing destination towards the desired domain or IP address - in this case, CFPlatform.xyz. Each line contains an IP, which are typically routers/switches within an ISP or backbone provider (global internet carrier), with the final destination being the IP CFPlatform.xyz resolved to. Within each line, the latency is displayed from your location to that hop and back, repeated until the destination is reached.

Common situations to look out for:

• * * * does not always indicate a problem on the hop/route - some routers/switches actively block requests (ICMP requests)
• A hop may have a sudden latency spike. This could be a handful of situations but typically because either:
  • The routing has navigated further than expected. For example if in the UK and you traceroute to a domain that’s also resolved to an IP in the UK, but you end up routing through New York, this is typically due to misconfigured or badly advertised routing. This can be either the ISP making a ‘cheaper’ routing decision or, more commonly, a routing misconfiguration. Depending on the latency swing, it's best to either contact ClearFront if the suspicion lies within their infrastructure or the offending ISP
  • One of the 3 latency values on the hop is high, for example, 3.605 ms 120.012ms 3.766ms, this could be network jitter or the switch itself may be overloaded. It can be difficult to judge; in the case that the latency is observed in CFPlatform.xyz hops, we can check the performance of our infrastructure directly, but are largely blind when the next route down has an issue like this. Latency variation is normal (typically less than 10ms)
• Sub-optimal routing with many hops - sometimes routes can be unwieldy and have tens of hops. This can introduce more potential points of failure as the packets traverse more networks and routers.