HP EVA P6000, User Manual

Page: 35 / 298

2 P6300/P6500 EVA operation
Best practices
For useful information on managing and configuring your storage system, see the HP P6300/P6500
Enterprise Virtual Array configuration best practices white paper available at:
http://h18006.www1.hp.com/storage/arraywhitepapers.html
Operating tips and information
Reserving adequate free space
To ensure efficient storage system operation, reserve some unallocated capacity, or free space, in
each disk group. The recommended amount of free space is influenced by your system configuration.
For guidance on how much free space to reserve, see the HP P6300/P6500 Enterprise Virtual
Array configuration best practices white paper. See
“Best practices” (page 35) .
Using SAS-midline disk drives
SAS-midline drives are designed for lower duty cycle applications such as near online data
replication for backup. Do not use these drives as a replacement for EVA's high performance,
standard duty cycle, Fibre Channel drives. This practice could shorten the life of the drive.
Failback preference setting for HSV controllers
Table 7 (page 35) describes the failback preference setting for the controllers.
Table 7 Failback preference settings
Behavior Point in time Setting
The units are alternately brought online to
Controller 1 or to Controller 2. At initial presentation No preference
If cache data for a LUN exists on a particular
controller, the unit will be brought online there. On dual boot or controller resynch
Otherwise, the units are alternately brought
online to Controller 1 or to Controller 2.
All LUNs are brought online to the surviving
controller. On controller failover
All LUNs remain on the surviving controller.
There is no failback except if a host moves the
LUN using SCSI commands. On controller failback
The units are brought online to Controller 1. At initial presentation Path A - Failover Only
If cache data for a LUN exists on a particular
controller, the unit will be brought online there. On dual boot or controller resynch
Otherwise, the units are brought online to
Controller 1.
All LUNs are brought online to the surviving
controller. On controller failover
All LUNs remain on the surviving controller.
There is no failback except if a host moves the
LUN using SCSI commands. On controller failback
The units are brought online to Controller 2. At initial presentation Path B - Failover Only
If cache data for a LUN exists on a particular
controller, the unit will be brought online there. On dual boot or controller resynch
Best practices 35