You can set a schedule to deduplicate a file system automatically by using the fsdedupadm setschedule command. You can specify two categories of schedule options: run periodicity, and type periodicity. The granularity of schedule is limited to the time of day and the day of the month. The fsdedupadm command applies any relevant File Change Log tunables when setting the schedule.

See Veritas File System File Change Log administrative interface.

You must enable deduplication on the file system before you can set a schedule.

See Enabling and disabling deduplication on a file system.

You can schedule the deduplication run every hour or every specified number of hours, and every day or every specified number of days. You can also schedule the actual deduplication run to occur each time, or every specified number of times that the scheduled time elapses. During times that deduplication does not occur, the deduplication run only updates the fingerprints in the database.

The schedule commands are not cumulative. If a deduplication schedule comes up while the previous deduplication process is running for any reason, the upcoming deduplication is discarded and an warning message displays.

You can remove a schedule by specifying an empty string enclosed by double quotes ("") for the schedule.

See the fsdedupadm(1M) manual page.

You must start the fsdedupschd daemon before scheduling the task:

# chkconfig --add fsdedupschd
# service fsdedupschd start

In the following example, deduplication for the file system /vx/fs1 will be done at midnight, every other day:

# fsdedupadm setschedule "0 */2" /vx/fs1

In the following example, deduplication for the file system /vx/fs1 will be done twice every day, once at midnight and once at noon:

# fsdedupadm setschedule "0,12 *" /vx/fs1

In the following example, deduplication for the file system /vx/fs1 will be done four times every day, but only the fourth deduplication run will actually deduplicate the file system. The other runs will do the scanning and processing. This option achieves load distribution not only in a system, but also across the cluster.

# fsdedupadm setschedule "0,6,12,18 * 4" /vx/fs1

The following example removes the deduplication schedule from the file system /vx/fs1:

# fsdedupadm setschedule "" /vx/fs1