Last updated

Linux and Unix wget command tutorial with examples

Tutorial on using wget, a Linux and UNIX command for downloading files from the Internet. Examples of downloading a single file, downloading multiple files, resuming downloads, throttling download speeds and mirroring a remote site.

Estimated reading time: 7 minutes

Table of contents

Terminal showing wget man page

What is the wget command?

The wget command is a command line utility for downloading files from the Internet. It supports downloading multiple files, downloading in the background, resuming downloads, limiting the bandwidth used for downloads and viewing headers. It can also be used for taking a mirror of a site and can be combined with other UNIX tools to find out things like broken links on a site.

How to download a file

To download a file with wget pass the resource your would like to download.

wget http://archlinux.mirrors.uk2.net/iso/2016.09.03/archlinux-2016.09.03-dual.iso
--2016-09-16 11:04:40--  http://archlinux.mirrors.uk2.net/iso/2016.09.03/archlinux-2016.09.03-dual.iso
Resolving archlinux.mirrors.uk2.net (archlinux.mirrors.uk2.net)... 83.170.94.3
Connecting to archlinux.mirrors.uk2.net (archlinux.mirrors.uk2.net)|83.170.94.3|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 792723456 (756M) [application/x-iso9660-image]
Saving to: ‘archlinux-2016.09.03-dual.iso’

archlinux-2016.09.03-dual.iso   2%[>                                                ]  18.48M  2.39MB/s    eta 5m 23s

The output of the command shows wget connecting to the remote server and the HTTP response. In this case we can see that the file is 758M and is a MIME type of application/x-iso9660-image. The file will be saved as archlinux-2016.09.03-dual.iso. Finally the standard output of wget provides a progress bar. This contains the following from left to right.

How to change the name of a downloaded file

To change the name of the file that is saved locally pass the -O option. This can be useful if saving a web page with query parameters. In the following example suppose that the URL https://petition.parliament.uk/petitions?page=2&state=all is to be downloaded.

wget "https://petition.parliament.uk/petitions?page=2&state=all"
--2016-09-16 11:15:26--  https://petition.parliament.uk/petitions?page=2&state=all
Resolving petition.parliament.uk (petition.parliament.uk)... 52.48.191.89, 54.72.19.76
Connecting to petition.parliament.uk (petition.parliament.uk)|52.48.191.89|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 25874 (25K) [text/html]
Saving to: ‘petitions?page=2&state=all’

petitions?page=2&state=all    100%[================================================>]  25.27K  --.-KB/s    in 0.02s

2016-09-16 11:15:27 (1.55 MB/s) - ‘petitions?page=2&state=all’ saved [25874/25874]

Using the ls command shows that the file has been saved as petitions?page=2&state=all. To specify a different filename the -O option may be used.

wget "https://petition.parliament.uk/petitions?page=2&state=all" -O petitions.html
--2016-09-16 11:18:04--  https://petition.parliament.uk/petitions?page=2&state=all
Resolving petition.parliament.uk (petition.parliament.uk)... 52.48.191.89, 54.72.19.76
Connecting to petition.parliament.uk (petition.parliament.uk)|52.48.191.89|:443... connected.
HTTP request sent, awaiting response... 200 OK
Length: 25874 (25K) [text/html]
Saving to: ‘petitions.html’

petitions.html                100%[================================================>]  25.27K  --.-KB/s    in 0.02s

2016-09-16 11:18:06 (1.38 MB/s) - ‘petitions.html’ saved [25874/25874]

How to turn off output

To turn off output use the -q option. This prevents wget from writing to standard output and makes it totally silent.

wget -q http://www.bbc.co.uk
ls
index.html

How to turn off verbose output

To turn off verbose output use the -nv option. This limits the output of wget but provides some useful information.

wget -nv http://www.bbc.co.uk
2016-09-16 11:23:31 URL:http://www.bbc.co.uk/ [172348/172348] -> "index.html" [1]

How to resume a download

To resume a download use the -c option. This makes wget for a file in the folder that the command was run from of the same name as the remote file. If there is a file then wget will start the download from the end of the local file. This can be useful if a remote server dropped a connection in the middle of a download or if your network dropped during a download. In the following example the download of the Arch Linux iso is resumed.

wget -c http://archlinux.mirrors.uk2.net/iso/2016.09.03/archlinux-2016.09.03-dual.iso
--2016-09-16 11:30:15--  http://archlinux.mirrors.uk2.net/iso/2016.09.03/archlinux-2016.09.03-dual.iso
Resolving archlinux.mirrors.uk2.net (archlinux.mirrors.uk2.net)... 83.170.94.3
Connecting to archlinux.mirrors.uk2.net (archlinux.mirrors.uk2.net)|83.170.94.3|:80... connected.
HTTP request sent, awaiting response... 206 Partial Content
Length: 792723456 (756M), 734083359 (700M) remaining [application/x-iso9660-image]
Saving to: ‘archlinux-2016.09.03-dual.iso’

archlinux-2016.09.03-dual.iso   7%[+++                                              ]  58.36M  2.02MB/s 

The download is resumed from the end of the local file.

How to download multiple URLs

To download multiple files at once pass the -i option and a file with a list of the URLs to be downloaded. URLs should be on a separate line. In the following example a listed of Linux ISOs is saved in a file called isos.txt.

http://archlinux.mirrors.uk2.net/iso/2016.09.03/archlinux-2016.09.03-dual.iso
https://www.mirrorservice.org/sites/cdimage.ubuntu.com/cdimage/releases/16.04/release/ubuntu-16.04.1-server-arm64.iso
http://mirror.ox.ac.uk/sites/mirror.centos.org/6.8/isos/x86_64/CentOS-6.8-x86_64-minimal.iso

To download these files in sequence pass the name of the file to the -i option.

wget -i isos.txt

How to throttle download speeds

To throttle download speeds use the --limit-rate option. This will limit the amount of bandwidth available to wget and can be useful to prevent wget consuming all the available bandwidth. In the following example the download is limited to 400k.

wget --limit-rate 400k http://archlinux.mirrors.uk2.net/iso/2016.09.03/archlinux-2016.09.03-dual.iso
--2016-09-16 12:05:12--  http://archlinux.mirrors.uk2.net/iso/2016.09.03/archlinux-2016.09.03-dual.iso
Resolving archlinux.mirrors.uk2.net (archlinux.mirrors.uk2.net)... 83.170.94.3
Connecting to archlinux.mirrors.uk2.net (archlinux.mirrors.uk2.net)|83.170.94.3|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 792723456 (756M) [application/x-iso9660-image]
Saving to: ‘archlinux-2016.09.03-dual.iso.6’

archlinux-2016.09.03-dual.iso   0%[                                                 ]   1.42M   400KB/s    eta 32m 12s

It is possible to see in the output the download is limited to 400KB/s.

How to download in the background

To download in the background use the background use the -b option. This can be useful in that no terminal is needed to run the download. The wget command will return the pid number of the process and also the name of the file that output will be logged to.

wget -b http://archlinux.mirrors.uk2.net/iso/2016.09.03/archlinux-2016.09.03-dual.iso
Continuing in background, pid 4135.
Output will be written to ‘wget-log’.

How to log output to a file

To direct wget output to a log file use the -o option and pass the name of a file.

wget -o iso.log http://archlinux.mirrors.uk2.net/iso/2016.09.03/archlinux-2016.09.03-dual.iso

To append output to a file use the -a option. If no file is present it will be created.

wget -a iso.log http://archlinux.mirrors.uk2.net/iso/2016.09.03/archlinux-2016.09.03-dual.iso

How to mirror a website

The wget command can mirror a remote website for local, offline browsing. It has many options for converting links and limiting downloads of certain file types.

wget -mk -w 20 http://www.example.com/

The options here are as follows

Once the command has finished a local copy of the remote site will be available.

To find broken links on a site wget can spider the site and present a log file that can be used to identify broken links.

wget -o wget.log -r -l 10 --spider http://example.com

The options in this command are as follows

The output file can then generate a list of unique broken links with the following command

grep -B 2 '404' wget.log | grep "http" | cut -d " " -f 4 | sort -u

How to set the user agent

To set the user agent pass the -U option. This allows an arbitrary string to be set for the user agent. This can be useful if the site you are trying to download blocks access to the wget user agent.

wget -U 'My-User-Agent' http://www.foo.com

A list of User Agents is available here.

How to view response headers

To view response headers pass the -S option. This will show the response headers from the server as well as downloading the file.

wget -S http://www.bbc.co.uk
--2016-09-16 14:20:52--  http://www.bbc.co.uk/
Resolving www.bbc.co.uk (www.bbc.co.uk)... 212.58.244.70, 212.58.246.94
Connecting to www.bbc.co.uk (www.bbc.co.uk)|212.58.244.70|:80... connected.
HTTP request sent, awaiting response...
  HTTP/1.1 200 OK
  Server: nginx
  Content-Type: text/html; charset=utf-8
  ...
Length: 171925 (168K) [text/html]
Saving to: ‘index.html’


index.html                    100%[================================================>] 167.90K  --.-KB/s    in 0.1s

2016-09-16 14:20:52 (1.44 MB/s) - ‘index.html’ saved [171925/171925]

To just view the headers and not download the file use the --spider option.

wget -S --spider http://www.bbc.co.uk
Spider mode enabled. Check if remote file exists.
--2016-09-16 14:23:42--  http://www.bbc.co.uk/
Resolving www.bbc.co.uk (www.bbc.co.uk)... 212.58.244.67, 212.58.246.91
Connecting to www.bbc.co.uk (www.bbc.co.uk)|212.58.244.67|:80... connected.
HTTP request sent, awaiting response...
  HTTP/1.1 200 OK
  Server: nginx
  Content-Type: text/html; charset=utf-8
  ...
Length: 171933 (168K) [text/html]
Remote file exists and could contain further links,
but recursion is disabled -- not retrieving.

Further reading

Have an update or suggestion for this article? You can edit it here and send me a pull request.

Tags

Recent Posts