Classic SysAdmin: 14 tail and head commands in Linux/Unix
The Linux Foundation | 16 April 2022
This is a classic article written by Surendra Anne from the Linux.com archives. For more great SysAdmin tips and techniques check out our Essentials of Linux System Administration course!
Many people know about cat command which is useful in displaying entire file content. But in some cases we have to print part of file. In today’s post we will be talking about head and tail commands, which are very useful when you want to view a certain part at the beginning or at the end of a file, specially when you are sure you want to ignore the rest of the file content.
let’s start with the tail command, and explore all of the features this handy command can provide and see how to use it best to suit your needs. After that we will show some options that you can do and can not do with the head command.
Linux Tail Command Syntax
tail [OPTION]... [FILE]...
Tail is a command which prints the last few number of lines (10 lines by default) of a certain file, then terminates.
Example 1: By default “tail” prints the last 10 lines of a file, then exits.
tail /path/to/file
Example :
# tail /var/log/messages Mar 20 12:42:22 hameda1d1c dhclient[4334]: DHCPREQUEST on eth0 to 255.255.255.255 port 67 (xid=0x280436dd) Mar 20 12:42:24 hameda1d1c avahi-daemon[2027]: Registering new address record for fe80::4639:c4ff:fe53:4908 on eth0.*. Mar 20 12:42:28 hameda1d1c dhclient[4334]: DHCPREQUEST on eth0 to 255.255.255.255 port 67 (xid=0x280436dd) Mar 20 12:42:28 hameda1d1c dhclient[4334]: DHCPACK from 10.76.198.1 (xid=0x280436dd) Mar 20 12:42:30 hameda1d1c avahi-daemon[2027]: Joining mDNS multicast group on interface eth0.IPv4 with address 10.76.199.87. Mar 20 12:42:30 hameda1d1c avahi-daemon[2027]: New relevant interface eth0.IPv4 for mDNS. Mar 20 12:42:30 hameda1d1c avahi-daemon[2027]: Registering new address record for 10.76.199.87 on eth0.IPv4. Mar 20 12:42:30 hameda1d1c NET[4385]: /sbin/dhclient-script : updated /etc/resolv.conf Mar 20 12:42:30 hameda1d1c dhclient[4334]: bound to 10.76.199.87 -- renewal in 74685 seconds. Mar 20 12:45:39 hameda1d1c kernel: usb 3-7: USB disconnect, device number 2
as you can see, this prints the last 10 lines of /var/log/messages.
Example 2: Now what about you are interested in just the last 3 lines of a file, or maybe interested in the last 15 lines of a file. this is when the -n option comes handy, to choose specific number of lines instead of the default 10.
tail -n <number_of_lines> /path/to/file
Example :
# tail -n 4 /etc/group vboxusers:x:491: avahi:x:70: mailnull:x:47: smmsp:x:51:
Example 3: We can even open multiple files using tail command with out need to execute multiple tail commands to view multiple files. Suppose if you want to see first two lines of a
tail -n <number of lines> <file1> <file2> <file3>
Example:
surendra@sanne-taggle:~/code/sh$ tail -n 2 99abc.txt startup_script.sh wifiactivate.sh ==> 99abc.txt <==
==> startup_script.sh <==sed -i 's/^.*PermitRootLogin.*$/PermitRootLogin yes/g' /etc/ssh/sshd_configservice sshd reload
==> wifiactivate.sh <==modprobe -rv iwlwifimodprobe -v iwlwifi 11n_disable=8
Example 4: Now this might be by far the most useful and commonly used option for tail command. Unlike the default behaviour which is to end after printing certain number of lines, the -f option “which stands for follow” will keep the stream going. It will start printing extra lines on to console added to the file after it is opened. This command will keep the file open to display updated changes to console until the user breaks the command.
tail -f /path/to/file
Example :
#service crond start ; tail -f /var/log/cron Starting crond: [ OK ] Mar 20 13:35:01 hameda1d1c CROND[5338]: (root) CMD (/root/mail/mailscript.sh) Mar 20 13:35:33 hameda1d1c crond[2354]: (CRON) INFO (Shutting down) Mar 20 13:35:50 hameda1d1c crond[5385]: (CRON) STARTUP (1.4.4) Mar 20 13:35:50 hameda1d1c crond[5385]: (CRON) INFO (RANDOM_DELAY will be scaled with factor 64% if used.) Mar 20 13:35:51 hameda1d1c crond[5385]: (CRON) INFO (running with inotify support) Mar 20 13:35:51 hameda1d1c crond[5385]: (CRON) INFO (@reboot jobs will be run at computer's startup.) Mar 20 13:36:01 hameda1d1c CROND[5390]: (root) CMD (/root/mail/mailscript.sh) Mar 20 13:36:12 hameda1d1c crond[5385]: (CRON) INFO (Shutting down) Mar 20 13:36:25 hameda1d1c crond[5436]: (CRON) STARTUP (1.4.4) Mar 20 13:36:25 hameda1d1c crond[5436]: (CRON) INFO (RANDOM_DELAY will be scaled with factor 85% if used.) ^C
As you can see in this example, I wanted to start the crond service, then watch the /var/log/cron log file as service starts. I used ; which a kind of command chaining in Linux in order to execute two commands in single line. I am not interested in just a few number of lines then exit, but moreover I am interested in keeping watching the whole log file till service starts, then break it with CTRL+C.
Example 5: The same tail -f command can be replicated using less command well. Once you open a file with less
less /path/to/filename
Once you open file, then press shift+f
Example:
Mar 21 08:25:01 sanne-taggle CRON[5553]: (root) CMD (command -v debian-sa1 > /dev/null && debian-sa1 1 1)Mar 21 08:27:24 sanne-taggle wpa_supplicant[807]: wlan0: WPA: Group rekeying completed with da:3c:69:04:b1:21 [GTK=CCMP]Mar 21 08:35:01 sanne-taggle CRON[5982]: (root) CMD (command -v debian-sa1 > /dev/null && debian-sa1 1 1)Waiting for data... (interrupt to abort)
In order to come out from update mode in less, you have to press ctrl+c and then press q for quit.
Example 6: We have other option -s which should always be used with -f” will determine the sleep interval, whereas tail -f will keep watching the file, the refresh rate is each 1 second, if you wish to control this, then you will have to use the -s option “sleep” and specify the sleep interval
tail -f -s <sleep interval in seconds> /path/to/file
Example :
# tail -f -s 5 /var/log/secureMar 20 12:43:27 sa su: pam_unix(su:session): session opened for user rabbitmq by (uid=0)Mar 20 12:43:27 sa su: pam_unix(su:session): session closed for user rabbitmqMar 20 12:43:27 sa su: pam_unix(su:session): session opened for user rabbitmq by (uid=0)Mar 20 12:43:28 sa su: pam_unix(su:session): session closed for user rabbitmqMar 20 12:43:28 sa su: pam_unix(su:session): session opened for user rabbitmq by (uid=0)
Example 7: As we seen in example 3, We can open more files using tail command. Even we can view 2 files at the same time growing using -f option as well. It will also print a header viewing which file is showing this output. the header line will be beginning with “==>”
tail /path/to/file1 /path/to/file2
Example:
# tail -f /var/log/secure /var/log/cron ==> /var/log/secure <== Mar 20 13:13:19 sa su: pam_unix(su:session): session opened for user rabbitmq by (uid=0) Mar 20 13:13:20 sa su: pam_unix(su:session): session closed for user rabbitmq Mar 20 13:13:25 sa su: pam_unix(su:session): session opened for user rabbitmq by (uid=0) Mar 20 13:13:26 sa su: pam_unix(su:session): session closed for user rabbitmq Mar 20 13:13:26 sa su: pam_unix(su:session): session opened for user rabbitmq by (uid=0) Mar 20 13:13:26 sa su: pam_unix(su:session): session closed for user rabbitmq Mar 20 13:13:26 sa su: pam_unix(su:session): session opened for user rabbitmq by (uid=0) Mar 20 13:13:27 sa su: pam_unix(su:session): session closed for user rabbitmq Mar 20 13:13:27 sa su: pam_unix(su:session): session opened for user rabbitmq by (uid=0) Mar 20 13:13:27 sa su: pam_unix(su:session): session closed for user rabbitmq ==> /var/log/cron <== Mar 20 13:00:02 sa CROND[19837]: (root) CMD (/usr/lib64/sa/sa1 1 1) Mar 20 13:01:01 sa CROND[20705]: (root) CMD (run-parts /etc/cron.hourly) Mar 20 13:01:01 sa run-parts(/etc/cron.hourly)[20705]: starting 0anacron Mar 20 13:01:01 sa run-parts(/etc/cron.hourly)[20714]: finished 0anacron Mar 20 13:01:01 sa run-parts(/etc/cron.hourly)[20705]: starting logrotate Mar 20 13:01:01 sa run-parts(/etc/cron.hourly)[20721]: finished logrotate Mar 20 13:02:01 sa CROND[21587]: (root) CMD (/etc/puppet/scripts/update-federation-links.py) Mar 20 13:10:01 sa CROND[28672]: (root) CMD (/usr/lib64/sa/sa1 1 1) Mar 20 13:13:00 sa crontab[31759]: (root) LIST (root) Mar 20 13:13:01 sa CROND[31817]: (root) CMD (/etc/puppet/scripts/update-federation-links.py) ^C
Example 8: If you want to remove this header, use the -q option for quiet mode.
Example :
# tail -fq /var/log/secure /var/log/cron Mar 20 13:13:19 sa su: pam_unix(su:session): session opened for user rabbitmq by (uid=0) Mar 20 13:13:20 sa su: pam_unix(su:session): session closed for user rabbitmq Mar 20 13:13:25 sa su: pam_unix(su:session): session opened for user rabbitmq by (uid=0) Mar 20 13:13:26 sa su: pam_unix(su:session): session closed for user rabbitmq Mar 20 13:13:26 sa su: pam_unix(su:session): session opened for user rabbitmq by (uid=0) Mar 20 13:13:26 sa su: pam_unix(su:session): session closed for user rabbitmq Mar 20 13:13:26 sa su: pam_unix(su:session): session opened for user rabbitmq by (uid=0) Mar 20 13:13:27 sa su: pam_unix(su:session): session closed for user rabbitmq Mar 20 13:13:27 sa su: pam_unix(su:session): session opened for user rabbitmq by (uid=0) Mar 20 13:13:27 sa su: pam_unix(su:session): session closed for user rabbitmq Mar 20 13:01:01 sa CROND[20705]: (root) CMD (run-parts /etc/cron.hourly) Mar 20 13:01:01 sa run-parts(/etc/cron.hourly)[20705]: starting 0anacron Mar 20 13:01:01 sa run-parts(/etc/cron.hourly)[20714]: finished 0anacron Mar 20 13:01:01 sa run-parts(/etc/cron.hourly)[20705]: starting logrotate Mar 20 13:01:01 sa run-parts(/etc/cron.hourly)[20721]: finished logrotate Mar 20 13:02:01 sa CROND[21587]: (root) CMD (/etc/puppet/scripts/update-federation-links.py) Mar 20 13:10:01 sa CROND[28672]: (root) CMD (/usr/lib64/sa/sa1 1 1) Mar 20 13:13:00 sa crontab[31759]: (root) LIST (root) Mar 20 13:13:01 sa CROND[31817]: (root) CMD (/etc/puppet/scripts/update-federation-links.py) Mar 20 13:20:01 sa CROND[8700]: (root) CMD (/usr/lib64/sa/sa1 1 1)
Example 9: Now what if I have a very huge /var/log/messages and I am only interested in the last certain number of bytes of data, the -c option can do this easily. observe the below example where I want to view only the last 500 bytes of data from /var/log/messages
tail -c <number of bytes> /path/to/file
Example :
# tail -c 500 /var/log/messages failed to connect to device: failed to connect to nws://admin@localhost:56025/?group=Administrators&;cert=%2Fvar%2Flib%2Fpuppet%2Fssl%2Fcerts%2Fc569b530-6b4f-4d72-8417-48a556bf55a5.pem&key=%2Fvar%2Flib%2Fpuppet%2Fssl%2Fprivate_keys%2Fc569b530-6b4f-4d72-8417-48a556bf55a5.pem Mar 20 13:27:07 sa collectd[2249]: nwservice.py: Upstart service in status (status:) does not match nwipdbextractor Mar 20 13:27:07 sa collectd[2249]: nwservice.py: Upstart service in status (status:) does not match nwbroker
Now, since we have been talking for a while about tail, lets talk about “head” command.
Head command in Linux
Head command will obviously on the contrary to tail, it will print the first 10 lines of the file. Till this part of the post, the head command will do pretty much the same as tail in all previous examples, with exception to the -f option, there is no -f option in head, which is very natural since files will always grow from the bottom.
Head Command Syntax In Linux
head [OPTION]... [FILE]...
Example 10: As mention earlier print first 10 lines.
# head /etc/passwd root:x:0:0:root:/root:/bin/bash bin:x:1:1:bin:/bin:/sbin/nologin daemon:x:2:2:daemon:/sbin:/sbin/nologin adm:x:3:4:adm:/var/adm:/sbin/nologin lp:x:4:7:lp:/var/spool/lpd:/sbin/nologin sync:x:5:0:sync:/sbin:/bin/sync shutdown:x:6:0:shutdown:/sbin:/sbin/shutdown halt:x:7:0:halt:/sbin:/sbin/halt mail:x:8:12:mail:/var/spool/mail:/sbin/nologin uucp:x:10:14:uucp:/var/spool/uucp:/sbin/nologin
Example 11: Print first two lines of a file.
# head -n 2 /etc/passwd root:x:0:0:root:/root:/bin/bash bin:x:1:1:bin:/bin:/sbin/nologin
Example 12: this option lets you print all lines starting from a line number you specify, unlike Example 11 which will show you the first number of lines you provided.
head -n <number of lines preceeded with "-"> /path/to/file
Example :
# head -n -27 /etc/passwd root:x:0:0:root:/root:/bin/bash bin:x:1:1:bin:/bin:/sbin/nologin daemon:x:2:2:daemon:/sbin:/sbin/nologin adm:x:3:4:adm:/var/adm:/sbin/nologin lp:x:4:7:lp:/var/spool/lpd:/sbin/nologin sync:x:5:0:sync:/sbin:/bin/sync
As you can notice, in this example, it printed all the lines starting after line 27.
Combine Head And Tail Command In Linux
Example 13: As tail and head commands print different parts of files in an effective way, we can combine these two to print some advanced filtering of file content. To print 15th line to 20th line in /etc/passwd file use below example.
head -n 20 /etc/passwd | tail -n 5
Output:
syslog:x:101:104::/home/syslog:/bin/falsemessagebus:x:102:106::/var/run/dbus:/bin/falseusbmux:x:103:46:usbmux daemon,,,:/home/usbmux:/bin/falsednsmasq:x:104:65534:dnsmasq,,,:/var/lib/misc:/bin/falsekernoops:x:106:65534:Kernel Oops Tracking Daemon,,,:/:/bin/false
Example 14: Many people do not suggest above method to print from one line to other line. The above example is to show how we can combine these things. If you really want to print a particular line, use sed command as shown below.
Example:
$ sed -n '5p' /etc/passwdsync:x:4:65534:sync:/bin:/bin/sync
Ready to continue your Linux journey? Check out our Essentials of Linux System Administration course!
Similar Articles
Browse Categories
2023 Compliance and Security Cloud Computing Open Source Projects Linux How-To 2024 Diversity & Inclusion LF Research Blog Open Source Best Practices Linux Foundation Newsletter 2022 Training and Certification Research Cross Technology Linux lf blog research report linux blog LFX cybersecurity project news software development AI Cloud Native Computing Foundation Legal OpenSearch Topic: Data Announcements Financial Services In the news Networking and Edge lf events Data Governance Energy Featured Events Industry: Finance Industry: Fintech Interoperability LF Energy Open Mainframe Open Models OpenChain System Administration This week at FINOS Topic: Open Source Development Topic: Security Topic: Sustainability Web Application & Development amazon web services aws brand perception cloud native cncf community tools confidential computing challenges developer needs eBPF emerging technologies generative AI human capital japan spotlight kernel lf projects license compliance maintainer openssf research survey sbom skills development tech talent techtalentsurvey updates