HP Matrix Operating Environment and HP Storage Provisioning Manager support for port groups

Technical white paper | Matrix OE and SPM
9
WWNs and the owning logical servers. The logical servers can be output to XML using the lsmutil command on the CMS
(Central Management Server). The command is:
C:\> lsmutil -list -ls -xml -file ls_list_out.xml
The above command will generate a file named ls_list_out.xml. Search and editing tools can be used to find the
server WWNs and identify the owning logical servers. The generated XML file contains Sever WWN structures such as the
following:
<LsaWWN>
<id>303</id>
<value>50:01:43:80:02:60:00:09</value>
<nodeValue>50:01:43:80:02:60:00:0B</nodeValue>
<oldValue>50:01:43:80:02:60:00:09</oldValue>
<allocatedValue>50:01:43:80:02:60:00:09</allocatedValue>
<owner>111</owner>
<wwnNumber>1</wwnNumber>
<state>4</state>
<type>0</type>
<timeRequested>0</timeRequested>
<physicalPort>0</physicalPort>
<inProfile>false</inProfile>
<inProfileOrder>777</inProfileOrder>
<connectionType>UNKNOWN</connectionType>
</LsaWWN>
The ‘value’ field (within the LsaWWN object) is the server WWN. The ‘owner’ field is the internal ID of the logical server that
owns this server WWN. Search for the logical server that owns this WWN by searching for that ID. The result will be a
LsaLogicalServer object with that ID, similar to the following:
<LsaLogicalServer>
<id>111</id>
<name>pg_sample_ls_1</name>
<description></description>
Whatever procedure is used, the result is a list of logical servers that need to be migrated to new controller ports, and
information on whether or not the volumes being accessed are redundant. Note that a redundant volume can be identified
by its two paths. (It will have both a primary and secondary path.) Depending on the method used above, this will show up
visually in the GUI or as multiple LsaSanVolumePath objects under the volume within the generated XML file.
Activities prior to re-zoning
With the list of logical servers in hand, it is time to create a plan to do the actual rezoning.
Note: The rezoning operations can take disks offline for a period of time, causing issues with boot and data volume access.
Previously, it was determined if the impacted volumes use redundant paths (multi-path). In this context, redundant paths
identify the volume as being accessible using two different initiator WWNs from the same server. Having multiple paths
allows one path to be used for volume access while the other path is being rezoned to new controller ports. Storage access
will automatically failover as needed to use the live path during this process. When rezoning the second path, volume
access will failover to the newly zoned first path. This allows rezoning to take place on a running server, making it the
easiest solution. Note that all of the steps outlined in the section Perform rezoning using the script should be followed for
one path before starting the rezoning of the second path.
If any of the volumes are not redundant, or if it can’t be determined that they are redundant, then it is necessary to
deactivate (shutdown) the server before rezoning. Arrangements must be made to avoid interrupting the services that
depend on that path. For example, migrate the VM guests providing the affected services to other hypervisors, or arrange
for a service interval in which the services that depend on the path can be halted, the Rezone script run, and the services can
be restarted.