Introduction
An important part of managing servers is monitoring network connectivity.
There are a few tools that are simple to use but valuable to know. In this guide, we will discuss how to use a tool called tracerouteto diagnose where a network problem may be occurring .
We will also look at a utility called mtrthat combines most of the functionality of ping and traceroute in a single interface .
How to Use Traceroute
tracerouteis a simple tool that shows the path to a remote server. This can be anything from a website you are trying to visit to a printer on your local network.
The tracerouteprogram is installed by default on almost every Linux distribution, so you do not need to install it.
To call it, we just need to provide a website or IP address that we want to explore:
traceroute google.com
You will get output similar to the following:
Output
traceroute to google.com (173.194.38.137), 30 hops max, 60 byte packets
1 192.241.160.253 (192.241.160.253) 0.564 ms 0.539 ms 0.525 ms
2 192.241.164.241 (192.241.164.241) 0.487 ms 0.435 ms 0.461 ms
3 xe-3-0-6.ar2.nyc3.us.nlayer.net (69.31.95.133) 1.801 ms 1.802 ms 1.762 ms
4 144.223.28.73 (144.223.28.73) 0.583 ms 0.562 ms 0.550 ms
5 144.232.1.21 (144.232.1.21) 1.044 ms 1.048 ms 1.036 ms
6 74.125.49.212 (74.125.49.212) 0.494 ms 0.688 ms 0.643 ms
7 209.85.248.180 (209.85.248.180) 0.650 ms 209.85.248.178 (209.85.248.178) 0.621 ms 0.625 ms
8 72.14.236.208 (72.14.236.208) 0.618 ms 72.14.236.206 (72.14.236.206) 0.898 ms 72.14.236.208 (72.14.236.208) 0.872 ms
9 72.14.239.93 (72.14.239.93) 7.478 ms 7.989 ms 7.466 ms
10 72.14.232.73 (72.14.232.73) 20.002 ms 19.969 ms 19.975 ms
11 209.85.248.228 (209.85.248.228) 30.490 ms 72.14.238.106 (72.14.238.106) 34.463 ms 209.85.248.228 (209.85.248.228) 30.707 ms
12 216.239.46.54 (216.239.46.54) 42.502 ms 42.507 ms 42.487 ms
13 216.239.46.159 (216.239.46.159) 76.578 ms 74.585 ms 74.617 ms
14 209.85.250.126 (209.85.250.126) 80.625 ms 80.584 ms 78.514 ms
15 72.14.238.131 (72.14.238.131) 80.287 ms 80.560 ms 78.842 ms
16 209.85.250.228 (209.85.250.228) 171.997 ms 173.668 ms 170.068 ms
17 66.249.94.93 (66.249.94.93) 238.133 ms 235.851 ms 235.479 ms
18 72.14.233.79 (72.14.233.79) 233.639 ms 239.147 ms 233.707 ms
19 sin04s01-in-f9.1e100.net (173.194.38.137) 236.241 ms 235.608 ms 236.843 ms
The first line tells us about the conditions under which traceroute is running:
Output
traceroute to google.com (173.194.38.137), 30 hops max, 60 byte packets
It gives the specified host, the IP address that DNS returned for that domain, the maximum number of hops to check, and the size of the packet to be used.
The maximum number of hops can be set with the -mflag . If the host you are trying to trace is more than 30 hops away, you may need to specify a larger value here. The maximum value you can set is 255.
traceroute -m 255 obiwan.scrye.net
You can set the size of the packet sent to each hop by giving an integer after the hostname:
traceroute google.com 70
You will see output like this:
Output
traceroute to google.com (173.194.38.128), 30 hops max, 70 byte packets
1 192.241.160.254 (192.241.160.254) 0.364 ms 0.330 ms 0.319 ms
2 192.241.164.237 (192.241.164.237) 0.284 ms 0.343 ms 0.321 ms
Each line after the first line represents a "hop", or intermediate host, that your traffic must pass through to reach the computer represented by the host you specified.
Each line has the following format:
Output
hop_number host_name (IP_address) packet_round_trip_times
Here is an example of a hop you might see:
Output
3 nyk-b6-link.telia.net (62.115.35.101) 0.311 ms 0.302 ms 0.293 ms
The meaning of each field is as follows:
- hop_number : A consecutive count of the degrees of separation the host has from your computer. Traffic to higher-numbered hosts has to pass through more computers to be routed.
- host_name : If available, this field contains the result of a reverse DNS lookup of the host's IP address. If no information is returned from the reverse DNS query, the IP address itself is given.
- IP_address : This field contains the IP address of this network hop.
- packet_round_trip_times : The rest of the line gives the round-trip times for a packet to the host and back again. By default, three packets are sent to each host, and each attempt is appended to the end of the line.
If you want to change the number of packets tested against each host, you can specify a number with the -qoption as follows:
traceroute -q1 google.com
If you want to skip the reverse DNS lookup to speed up the trace, you can pass the -nflag :
traceroute -n google.com
You will get output like this:
Output
traceroute to google.com (74.125.235.7), 30 hops max, 60 byte packets
1 192.241.160.253 0.626 ms 0.598 ms 0.588 ms
2 192.241.164.241 2.821 ms 2.743 ms 2.819 ms
3 69.31.95.133 1.470 ms 1.473 ms 1.525 ms
If your trace turns into some asterisks (*), there is a problem with the route to the host.
Output
...
15 209.85.248.220 (209.85.248.220) 121.809 ms 72.14.239.12 (72.14.239.12) 76.941 ms 209.85.248.220 (209.85.248.220) 78.946 ms
16 72.14.239.247 (72.14.239.247) 101.001 ms 92.478 ms 92.448 ms
17 * * 209.85.250.124 (209.85.250.124) 175.083 ms
18 * * *
19 * * *
What Does a Route Problem Mean?
If your traceroute attempt stops at a particular hop or node and cannot find a path to the host, you have a problem.
The hop where the route does not come back may be the location of the networking problem, but it is not always that easy to diagnose.
Because each ping represents a round-trip packet, and because packets often take different paths in each direction, it may point to a problem on an entirely different, possibly closer, route.
The problem may also be on the hop immediately after the last one you see. If you cannot get a return trace path from that particular hop, it is difficult to diagnose the exact location of the problem. This is usually not possible outside your own network.
How to Use MTR
A dynamic alternative to the traceroute program is mtr. Combining the functionality of ping and traceroute, mtr lets you continuously poll a remote server and see how latency and performance change over time.
Unlike traceroute, mtr is not installed by default on most systems. You can get it by typing the following commands.
Ubuntu / Debian:
sudo apt-get install mtr
CentOS / Fedora:
yum install mtr
Arch:
pacman -S mtr
Once installed, you can call it by typing:
mtr google.com
You will get output like this:
Output
My traceroute [v0.80]
traceroute (0.0.0.0) Tue Oct 22 20:39:42 2013
Resolver: Received error response 2. (server failure)er of fields q
uit Packets Pings
Host Loss% Snt Last Avg Best Wrst StDev
1. 192.241.160.253 0.0% 371 0.4 0.6 0.1 14.3 1.0
2. 192.241.164.241 0.0% 371 7.4 2.5 0.1 37.5 4.8
3. xe-3-0-6.ar2.nyc3.us. 2.7% 371 3.6 2.6 1.1 5.5 1.1
4. sl-gw50-nyc-.sprintli 0.0% 371 0.7 5.0 0.1 82.3 13.1
Although the output looks similar, its biggest advantage over traceroute is that the output is continuously updated. This lets you accumulate trends and averages, and also see how network performance changes over time.
If you have run a traceroute, even in a situation where the route suffers from intermittent packet loss, the packets sent to each hop are likely to make the trip uneventfully. The mtr utility lets you monitor this situation by collecting data over a wider time span.
It is also possible to run mtr with the --reportoption, which returns the results of sending 10 packets to each hop .
mtr --report google.com
The report looks like this:
Output
HOST: traceroute Loss% Snt Last Avg Best Wrst StDev
1.|-- 192.241.160.254 0.0% 10 1.5 0.9 0.4 1.5 0.4
2.|-- 192.241.164.237 0.0% 10 0.6 0.9 0.4 2.7 0.7
3.|-- nyk-b6-link.telia.net 0.0% 10 0.5 0.5 0.2 0.7 0.2
4.|-- nyk-bb2-link.telia.net 0.0% 10 67.5 18.5 0.8 87.3 31.8
This can be useful when you do not want to measure in real time but want a broader range of data than traceroute provides.