Hitachi Dynamic Link Manager Software Users Guide for AIX (6.6) (T5208-96010, April 2011)

4. HDLM Operation
209
procedure when all the following conditions are satisfied
#
, the volume group
cannot be activated after the host is restarted:
• The host was shut down while a volume group was active, and the volume
group consisted of hdisks that had paths going through the HBAs to be
replaced.
• The hdisk reservation policy for the hdisks that make up the volume group
has been set to
PR_exclusive.
#
If you replace some HBAs for the volume group, after the host is restarted,
the volume group is activated and the paths going through the new HBAs are
added. However, the paths that go through the old HBAs remain defined, so
delete such paths as necessary.
To activate a volume group, perform the following:
• For a volume group comprising rootvg
Clear the reservation for the applicable LUs from the storage system or other
servers.
• For a volume group other than rootvg
1. If LUN security has been set up for the storage system, add the WWNs of the
new HBAs to the LUN security.
2. Start the host.
3. Execute the following command to delete the HDLM devices on the paths
that go through the old HBAs:
# rmdev -dl hdisk-name
n indicates the instance number of the HDLM device.
4. If LUN security has been set up for the storage system, delete the WWNs of
the old HBAs from the LUN security.
5. Execute the following command to reconfigure the devices:
# cfgmgr -l fcsn
Specify the device name of the PCI slot (
fcs device) where each HBA has
been replaced. n indicates the instance number of the
fcs device.
6. Execute the following command to check path information:
# /usr/DynamicLinkManager/bin/dlnkmgr view -path