Hitachi 9900 Series User Manual
Page 36
There are four basic types of high-availability middleware that can reduce downtime in
the event of a data path or host failure. The Lightning 9900
™
supports each classification of
middleware:
1.
Alternate Pathing middleware switches the I/O load in the event of
path failure
This type of middleware automatically switches the I/O load on a failed primary
path to an alternate path on the same host system. The Lightning 9900 support
alternate pathing for AIX
®
4.2 and above, Compaq True64
™
UNIX
®
, HP
®
10.01
and through PVlink, Windows NT/2000
®
and Sun Solaris
™
2.5 and Sequent
DYNIX/ptx
®
. In addition, VERITAS
®
Dynamic Path Management
™
(DPM) is
supported.
2 . Hitachi Dynamic Link Manager
™
provides path failover and load
balancing
Hitachi Dynamic Link Manager
™
is a family of Hitachi-provided middleware
software utilities that are server-based as shown in Figure 22. Hitachi
Dynamic Link Manager
™
enhances the availability of RAID systems by providing
automatic error recovery and path failover from server-to-RAID connection
failures. Hitachi Dynamic Link Manager
™
provides load balancing in addition to
path failover by re-directing I/O activity to the least busy path using complex
algorithms.
Just because a system is RAID-protected doesn’t mean it is protected against
connection bus failures, which is why Hitachi Data Link Manager
™
is required for
true non-stop operations. This product allows systems administrators to take
advantage of the multiple paths on a Lightning 9900
™
by adding redundant
connections between application servers and RAID systems. Hitachi Dynamic
Link Manager
™
therefore provides increased reliability and performance.
Supported platforms include AIX
®
, Sun Solaris
™
, Windows NT
®
, and Windows
2000
®
.
3.
Host Failover
Host failover software supports a cluster of host processors in which one of the
hosts automatically takes over the workload of any failed host in the cluster. This
“take-over” includes the reassignment of networks and peripherals, as well as the
restarting of applications. Host clustering can also be used to create fault-tolerant
work loads and scale processor capability while sharing network and disk
resources.
The Lightning 9900
™
supports all major open systems clustering schemes
including: Compaq TruCluster
®
, HP
®
MC Service Guard, HP
®
MC Lock
Manager, IBM
®
RS6000 and SP HACMP, Windows NT/2000
®
, Microsoft
®
Cluster Server, NCR
®
UNIX
®
SVR4 Lifekeeper, IBM DYNIX/ptx ATAP
®
, and
VERITAS ClusterServer
™
for Sun Solaris
™
.
Alternate pathing
automatically switches
the I/O load on a failed
primary path.
Hitachi Dynamic Link
Manager
™
provides load
balancing in addition to
path failover.
One of the hosts
automatically takes
over the workload of any
failed host in the cluster.
27
Hitachi High Availability Software Improves Application Service Levels