Best practices, Select an optimum recurrence schedule, Example configuration – HP LeftHand P4000 Virtual SAN Appliance Software User Manual
Page 40: Using remote copy for off-site backup and recovery, Achieving off-site backup
Best practices
Select an optimum recurrence schedule.
Select a recurrence schedule for remote snapshots that minimizes the potential for data loss. Any
data written to the primary volume since the most recent remote snapshot was created will be
unavailable if the primary volume is unavailable. Consider how much data you are willing to lose
in the event of an emergency and set the recurrence for creating remote snapshots accordingly.
If you do not want a large number of remote snapshots to accumulate on your remote volume, you
can use several timetables for scheduled remote snapshots, each with different retention policies.
For example, suppose you want to create remote snapshots every four hours to ensure that no more
than four hours worth of data is lost in an emergency. In addition, you want to retain one week's
worth of remote snapshots. Retaining four-hour snapshots for one week can result in the accumulation
of over 40 remote snapshots. Another approach would be to create two remote snapshot schedules
for the volume:
•
One schedule to create remote snapshots every four hours, but only retain the most recent six
remote snapshots. This ensures that you do not lose more than four hours worth of data in an
emergency.
•
A second schedule to create remote snapshots every 24 hours and retain 7 remote snapshots.
Use remote snapshots in conjunction with local, synchronous volume replication
To prevent data loss, reinforce Remote Copy with Network RAID within the cluster of storage
systems at the primary geographic site. With Network RAID configured, a single storage system
can be offline, and your primary volume remains intact.
At the backup location, you can also use synchronous replication to protect your remote volume
against storage system failure.
Example configuration
•
In the production location, create a cluster of three storage systems, all with managers.
•
Create volumes on the cluster, and set the data protection level to Network RAID-10.
•
Create a schedule for making remote snapshots of the primary volume. Set the recurrence to
every four hours, and retention of remote snapshots to two days.
NOTE:
You can use the same data protection levels on the remote volume as well. However, this
data protection level is configured independently of the data protection level that is configured on
the primary volume.
If one of the storage systems in the primary location fails, the primary volume will still be available.
If all of the storage systems fail, or if the application server fails, then you can recover data from
the remote snapshots or tape backups in the off-site location.
Using Remote Copy for off-site backup and recovery
For backup and recovery systems, Remote Copy can eliminate the backup window on an application
server. Using iSCSI command line interface commands and scripts, configure the iSCSI initiator to
mount remote snapshots on a local or remote backup server, and then back up the remote snapshot
from the backup server. The remote snapshot is available if the primary volume fails.
Achieving off-site backup
Rather than creating tape backups and then transporting them to a secure off-site location, you
can use Remote Copy to create remote snapshots in an off-site location. Then, optionally, you can
create tape backups at the off-site location.
40
Sample Remote Copy configurations