4.0

You cannot designate a third site as a recovery site for one that is already paired with another site. If you want
to use SRM to provide business continuity and disaster recovery services at a recovery site, you must configure
that site as a protected site that uses its own array managers to replicate data to the other member of the site
pair. After site pairing is complete, configuring bidirectional operation requires you to follow the same site
configuration procedures that are required for unidirectional operation, but you must do so for each site in
each capacity. At recovery site that has not been configured for bidirectional operation, items that must be
configured at a protected site remain unconfigured:
n
Array Managers and Inventory Mappings are always listed as Not Configured.
n
Protection Groups are listed as No Groups Created.
How Changes to Virtual Machine Storage Affect Protection
When you edit the properties of a virtual machine to add or change storage devices (such as hard disks or DVD
drives) you can affect the protection of that machine if you connect it to a device that has storage on a datastore
that is not replicated, or that is protected by a different protection group.
n
If the new device is created on a replicated datastore that is not protected (not part of any protection group),
the datastore is added to the virtual machine's protected datastore group and the virtual machine's
protection is unaffected.
n
If the new device is created on a replicated datastore that is protected by a different protection group, the
virtual machine's protection is invalidated. If the new device is created on a nonreplicated datastore, the
virtual machine's protection is invalidated.
If you use Storage VMotion to move a virtual machine to a nonreplicated datastore, or to a replicated datastore
on an array that SRM has not been configured to manage (through an SRA), the virtual machine's protection
is invalidated.
How Site Recovery Manager Computes Datastore Groups
The composition of a datastore group is determined by the set of virtual machines that have files on the
datastores in the group, and by the devices on which those files are stored.
Virtual machine files are located on one or more vSphere datastores. Each datastore consists of one or more
extents. Each extent corresponds to a single partition of a device on a storage array. Array replication is
configured on a per-device basis; most arrays include some devices that are not replicated. SRM must ensure
that all devices containing protected virtual machine files are replicated. During a recovery or test, SRM must
failover all such devices together.
To solve this problem, SRM aggregates datastores into datastore groups. A datastore group consists of the
minimal set of devices required to ensure that if any of a virtual machine's files is stored on a device in the
group, all of the virtual machine's files are stored on devices that are part of the same group. For example, if a
virtual machine has disks on two different datastores, then both datastores must be combined into a datastore
group. Conditions that can cause datastores to be combined into a datastore group include:
n
A virtual machine has files on two different datastores.
n
Two virtual machines share an RDM device on a SAN array.
n
Two datastores span extents corresponding to different partitions of the same device.
n
A single datastore spans two extents corresponding to partitions of two different devices.
n
Multiple devices belong to a consistency group defined on the storage array.
SRM computes datastore groups when you first configure your array managers. After that, the computation
executes every time that a virtual machine is added to or removed from a datastore that is part of a group.
Chapter 1 Administering VMware vCenter Site Recovery Manager
VMware, Inc. 9