HP EVA Cluster Extension Software Administrator Guide (5697-2445, July 2013)

HP P6000 Command View uses
HP P6000 Command View is used to create vdisks and configure the HP EVA storage system.
Using HP P6000 Command View, you can:
• Create EVA Continuous Access DR groups.
• Use Replication Solutions Manager or P6000 Command View to view a graphical
representation of a DR group setup.
• Use Replication Solutions Manager or P6000 Command View to manually switch source and
destination roles of DR group members.
SSSU considerations
You must use the SSSU to set the access mode of the destination vdisk to read-only. You can also
use SSSU to:
• Create vdisks and DR groups in a convenient automated fashion by using custom scripts.
• Capture your HP EVA storage system configuration setup.
IP network considerations
Because P6000 Command View and HP EVA SMI-S are essential resources for EVA Cluster
Extension, HP recommends that you provide reliable network connections for
cluster-node-to-management server communications. HP also recommends that you use the redundant
network connections of the management server for alternative access.
SAN fabric considerations
The management servers use the SAN to communicate with the EVA storage system. Make sure
that the management servers are configured to access both fabrics and EVA arrays in the EVA
Continuous Access configuration.
Cluster Shared Volume for Windows Server 2012
The Cluster Shared Volume feature of Microsoft Failover Cluster allows all the nodes in a cluster
to directly access the same volume, without changing ownership of the disk resource. Because of
this feature, all the nodes in a cluster can use the same volume to simultaneously host active virtual
machines. CSV manages storage access differently compared to regular clustered disks. A CSV
volume is a shared disk containing NTFS partitions. CSV allows you to store multiple VHDs on a
single LUN and run the associated VMs on any cluster node in read-write mode. This makes the
failover operation quicker. CSV also enables quicker live migration that allows you to move a
running VM from one node to another node with zero downtime.
As disk ownership does not change when a VM moves from one to another node, the process
becomes quicker and safer. This allows the client nodes to stay connected while the VM is moved
to another node in the cluster. The CSV disk resource that is owned by a cluster node is known as
a coordinator node. The VMs hosted on the CSV disk can be owned by either the coordinator
node or any other node in the cluster. When a disk is in the available storage group, it can be
converted to a CSV. When a CSV disk is created in the Failover Cluster, a role is internally created
by the failover cluster which is hidden, and is visible through only the cluster.exe command
output or through the PowerShell Cmdlet output, and not in the Failover Cluster GUI. Only CSV
disk is shown on the Failover Cluster GUI. The name of the CSV role is in the form of a GUID. For
every highly available VM that is created on the CSV, a VM role is created in the Failover Cluster
that is independent of the CSV role. Figure 5 (page 16) shows two virtual machines created on a
single CSV Disk. For each virtual machine, a role is created.
In a multi-site storage array replicated environment, CSV disks are replicated to the remote
datacenter using the array based remote replication feature. VMs residing in the CSV disk can run
on any cluster nodes (physical servers). In case of a disaster in the primary datacenter, CSV disks
Planning for EVA Cluster Extension 15