Extended Copy Service
The Extended Copy Service feature of VxVM automatically works in tandem with the extended copy engines from array vendors. When VxVM detects that the source and destination devices are enabled for extended copy, VxVM automatically off loads copy requests to the array's copy manager.
The benefits of VxVM Extended Copy Service are:
- Non-disruptive copy operations from disk to disk. The host server remains online during the copy and the data being copied remains accessible to the server.
- Server-free copy operation. The copy operation is done between the array subsystem and the target disk. The data copy operation does not use any CPU or I/O resources on the host server.
To see whether the Extended Copy Service feature is enabled on a disk, use the vxprint command as shown in the following example. The feature is enabled if an ecopy_enabled entry appears in the flags line.
# vxprint -l disk03
Disk group: rootdg
Disk: disk03
info: diskid=1234567890.1234.tungsten
assoc: device=sde
flags: autoconfig ecopy_enabled
device: pubpath=/dev/vx/dmp/sde
devinfo: publen=35354136 privlen=9167
Note
If required, you can use the -o noecopy option to turn off Extended Copy Service for each invocation of the vxplex att, cp, mv and snapstart commands, and the vxsd mv command.
Enabling a Disk for Ecopy Operation
To enable a disk for ecopy operation:
-
Install the Hardware Assisted copy license
-
Enable the Extended copy features in the array. (This procedure is vendor specific.)
-
Install the vendor ASL that supports the ecopy feature. Contact VITA@veritas.com for vendor ASL information.
-
Shutdown and reboot the system.
Extended Copy Service Support for Hitachi 9900 and 9900V Arrays
To implement extended copy for the Hitachi 9900 and 9900V arrays, you must create the two files, user_pwwn_file and user_luid_file, that identify the Hitachi disk identification information. To create these files, run the vxwwnluid utility that is located in /etc/vx/diag.d. There are no parameters for the vxwwnluid utility, and the command must be executed as root.
- /etc/vx/user_pwwn_file
This file contains the disk access name and the port world-wide-name (pwwn) for each disk in the array. For the hitachi arrays, both the source and the destination devices must have entries in the this file. The information for each disk in the array is defined on a single line. The disk access name and PWWN are separated by a single tab character.
The following are sample entries from the /etc/vx/user_pwwn_file file:
sde 50060e800404040b
sdf 50060e800404040b
sdg 50060e800404040b
- /etc/vx/user_luid_file
This file contains the disk access names and their corresponding LUN numbers in the array. The information for each disk in the array is defined on a single line. The disk access name and the LUN are separated by a single tab character.
The following are sample entries from the /etc/vx/user_luid_file file:
sde 1
sdf 2
sdg 1
|
