HP LeftHand P4000 SAN Solutions - Manual (Page 111)

Volume-level data protection
Use a data protection level greater than Network RAID-0 to ensure optimum data availability if a
storage system fails. For information about data protection, see
“Planning data protection”
(page 142) .
Volume access
Use iSCSI load balancing to ensure better performance and better utilization of cluster resources.
For more information about iSCSI load balancing, see
“iSCSI load balancing” (page 241) .
Systems running managers
Use the recommended number and type of managers to ensure optimum availability of your
management group and volumes. Configurations with 3 or 5 managers are recommended for most
single site installations. Three or five storage systems running managers is optimum. For fewer
storage systems, use a Failover Manager as the third manager if possible. The Virtual Manager is
also available for specific configurations. For a detailed discussion of managers and quorum, see
“Managers and quorum” (page 119) .
NOTE: This Best Practice appears in the Best Practice Summary for releases 9.5 and earlier.
Beginning with release 10.0, managers and quorum configuration issues are tracked using alarms
and events notifications.
Network bonding
Bonding the available NIC cards in each storage system improves SAN performance and reliability.
In most instances, Adaptive Load Balancing is the recommended bond. See
“Best practices”
(page 53) .
Network bond consistency
To ensure consistent failover characteristics and traffic distribution, use the same network bond
type in all the storage systems in a cluster. Inconsistent network bonds in storage systems in the
same cluster may cause inconsistent traffic across NICs, and unexpected failover behavior depending
on which system in the cluster fails. For more information, see
“Configuring network interface
bonds” (page 52) .
Network flow control consistency
Set the network flow control settings the same on every storage system in the cluster. Inconsistent
network flow control settings in storage systems in the same cluster may impact the connectivity
between the systems. For more information, see
“Changing NIC flow control” (page 50) .
Network frame size consistency
Network frame size affects management data traffic and replication. Use the same network frame
size in all storage systems in a cluster to ensure consistent data traffic and replication. Inconsistent
network frame size may impact data traffic and replication in the cluster. For more information,
see
“Changing NIC frame size” (page 49) .
Management group maintenance tasks
When you have an established management group, you may need to perform maintenance activities
on the group.
Logging in to a management group
You must log in to a management group to administer the functions of that group.
Management group maintenance tasks 111