Refer to the exhibit.
The customer wants to implement an HP StoreVirtual solution for the environment shown. The
storage requirements dictate that the solution must have four nodes per site. What needs to be
considered when designing the solution?

A.
the latency on the link

B.
a failover manager required on both sites

C.
200 MB/s bandwidth per storage nodes pair required

D.
the type of data being replicated

Explanation:
Best practices

• Use a single IP subnet with Layer 2 switching
• Configure redundant network paths between sites to help minimize link failure in a two-site
configuration. In multi-site cluster with three sites, HP recommends a triangle network design, not
a V-design to ensure high availability.
• For 1GbE installations allocate 50 MB/s per storage node in the sites between data centers; for
environments with 10GbE storage connectivity, reserve 250 MB/s. For campus cluster
deployments with Fibre Channel connectivity on HP StoreVirtual, plan for 500 MB/s per storage
node in the sites (only two sites supported).
• In the example (see Figure 7) above with 1GbE storage connectivity and two storage nodes in
each site, designate 50 MB/s x 2 = 100 MB/s of bandwidth between the sites (at least one 1GbE
link; two for redundancy).
• HP StoreVirtual Fail-over Manager is not in the data path and has to be connected using at least
Fast Ethernet (100 Mb/s) with a latency not exceeding 50 ms.
Network latency
High network latency can be the primary cause of slow I/O performance, or worse, iSCSI target
disconnects. It is important to keep round trip network latency on your storage subnet below two
milliseconds. Many factors can contribute to increasing network latency, but two are most
common:
• Distance between storage cluster nodes
• Router hops between storage cluster nodes
Configuring a multi-site cluster on a single IP subnet with Layer 2 switching will help to lower the
network latency between storage cluster nodes.