Building a Disaster-proof Data Center with HP Serviceguard for Linux, June 2007

13
Attributes Extended Distance Cluster CLX
Key Benefit
Excellent in “normal” operations, and partial
failure. Since all hosts have access to both
disks, in a failure where the node is running
and the application is up, but the disk
becomes unavailable, no failover occurs. The
node will access the remote disk to continue
processing.
Two significant benefits: • Provides maximum
data protection. State of the data is determined
before application is started. If necessary, data
resynchronization is performed before
application is brought up. • Better performance
than Extended Distance Cluster for
resynchronization, as replication is done by
storage subsystem (no impact to host).
Key Limitation
No ability to check the state of the data before
starting up the application. If the volume group
(vg) can be activated, the application will be
started. If mirrors are split or multiple paths to
storage are down, as long as the vg can be
activated, the application will be started. Data
resynchronization does not have a big impact
on system performance. However, the
performance varies depending on the number
of times data resynchronization occurs. In the
case of MD, data resynchronization is done
one disk at a time, using about 10% of the
available CPU time and taking longer to
resynchronize multiple LUNs. The amount of
CPU time used is a configurable MD
parameter.
Specialized storage required. Currently, XP
with Continuous Access, and EVA with
Continuous Access are supported.
Maximum Distance 100 Kilometers
Shortest of the distances between: • Cluster
network latency (not to exceed 200 ms). •
Data Replication Max Distance. • DWDM
provider max distance.
Data Replication
mechanism
Host-based, through MD. Replication can
affect performance (writes are synchronous).
Resynchronization can impact performance.
(Complete resynchronization is required in
many scenarios that have multiple failures.)
Array-based, through Continuous Access XP or
Continuous Access EVA. Replication and
resynchronization performed by the storage
subsystem, so the host does not experience a
performance hit. Incremental
resynchronizations are done, based on bitmap,
minimizing the need for full re-syncs.
Application Failover
type
Automatic (no manual intervention required). Automatic (no manual intervention required).
Access Mode for a
package
Active/Standby Active/Standby
Client Transparency
Client detects the lost connection. You must
reconnect once the application is recovered at
second site.
Client detects the lost connection. You must
reconnect once the application is recovered at
second site.
Maximum Cluster
Size Allowed
4 nodes 2 to 16 nodes
Storage
Identical storage is not required (replication is
host-based with MD mirroring).
Identical Storage is required.
Data Replication Link Dark Fiber
Dark Fiber Continuous Access over IP
Continuous Access over ATM
Cluster Network Single or multiple IP subnet Single or multiple IP subnet
DTS Software/
Licenses Required
Serviceguard for Linux + XDC Serviceguard for Linux + CLX XP or CLX EVA