You can execute batch jobs in UNIX / Linux using any one of the three commands — at, batch or cron.
In this article, let us review how to schedule a job, view a job, and delete a job using at command.
You can schedule an at job in two different ways:
- Schedule the job to be executed at a specific time. For example, July 3rd, 10AM
- Schedule the job to be executed in relative time from now. For example, 5 hours from now.
1. Schedule an at job using specific date and time
$ at time date
For example, to schedule a job at 11 am on May 20, use the following at command.
$ at 11 am may 20
2. Schedule an at job using relative time
You can schedule a job to be executed using relative time from now.
$ at now + COUNT UNIT
For example, following job will be execute 1 minute from now.
$ at now + 1 min
The above example will read the commands from stdin, and it will execute the job after a minute. When you give something wrong in time format, you will get the error ‘Garbled time‘.
You can schedule a background job for 1 hour from now, (or) 1 day from now using the following at command:
$ at now + 1 hour $ at now + 1 day
Similar to at command, you can also use crontab to execute jobs at a scheduled time. Refer to our earlier 15 cron command examples article.
3. View all the scheduled at jobs using atq
You can use atq command (or at -l), to display all the at command jobs that are scheduled or currently running.
The following atq command will list all the pending at jobs. The first number shown is the Job number, followed by the time in which the process is to be executed, and the user name.
$ atq 4 2010-04-20 11:00 a sathiya
4. Remove/Delete a scheduled at job using atrm
You can use atrm command (or at -d), to delete a particular job. For example, to delete the job number 4, use the following atrm command.
$ atrm 4
5. Execute a job only when system load average is < 1.5 using batch command
You can schedule a job using batch command, which will prompt for command input, which will be executed when the system load average is less than 1.5.
At the successful completion of input, you will get job number. For listing and removing batch jobs you can use the at commands explained above.
6. Schedule at jobs from file using -f option
First create a text file that contains all the commands, or shell-scripts that you would like to be executed in the background using at command.
$ cat myjobs.txt /home/sathiya/calculate-space.sh /path/to/a/shell-script /path/to/any/command/or/script
Using the -f option, you can make the at command to get the input from the file instead of stdin.
Following at command will execute all the jobs from the myjobs.txt 1 hour from now.
$ at -f myjobs.txt now + 1 hour
7. Allowing and Denying certain users from using at jobs
System administrator can control who can schedule an at job and who cannot using at.allow and at.deny files.
First, system checks for at.allow file. If at.allow exists, only the usernames specified in the at.allow file are allowed to use at command.
Next, (if at.allow doesn’t exist), system checks for at.deny file. If at.deny exist, the usernames specified in the at.deny file are not allowed to use the at command.
By default, most systems uses at.deny file to stop certain users from using the at command, such as www-data, guest, backup, man user.
8. Execute at command like nohup
Similar to the nohup command we discussed earlier, you can execute a command (or shell script) on the remote server using the at command and logout from the server.
$ at -f myjob now + 1 min $ exit
Note: myjob will still be running even after you exit out of the server.
9. Additional at command time formats
You can use any one of the following at command date time formats:
$ at 10 am tomorrow $ at 11:00 next month $ at 22:00 today $ at now + 1 week $ at noon