Implement high-availability solutions with HP Instant Capacity - easily and effectively

8
HP TiCAP and high availability
You can use TiCAP to provide cost-effective high-availability for any of the models covered in the “HP Serviceguard
failover models” section. To achieve these high-availability solutions, you must first configure iCAP processors on the
cluster nodes. These processors can then be activated temporarily in failover situations, using additional incremental
purchases of TiCAP. This solution provides a less costly solution because you do not pay full price for the iCAP
processors up front. The solution is cost-effective, because you only pay for extra capacity if a failover situation occurs.
Then, you start to consume your TiCAP balance.
When evaluating a high-availability solution using TiCAP, you should consider the following:
• TiCAP is typically used only when you do not have any core usage rights to migrate from other partitions of a server or
from other GiCAP group members (if any).
• TiCAP can be used only to activate cores. TiCAP cannot be used for additional memory resources.
• TiCAP can be used in any of the Serviceguard failover models. (However, other iCAP solutions may be more
cost-effective, particularly for the rotating standby model where continuous TiCAP consumption might result when
the standby becomes the active node.)
• TiCAP is a flexible and open-ended solution for temporary activation of processor cores.
• TiCAP is used when you can comfortably predict the cost and duration of a failover situation, because it requires
additional incremental purchases of capacity.
• TiCAP reserves can be maintained for emergency use while you arrange for long-term solutions.
Example: HP TiCAP manual failover
A server called Sys1 runs the primary applications for a particular site. You purchase it as a system with all usage rights
(no iCAP components). You equip Sys1 with eight active processor cores, each on two blades.
At the same time, you purchase server Sys2 with similar characteristics, except that all but four processor cores are
purchased as iCAP processors. (At least one processor core must be active for each hard partition [nPartition]. The
assumption is that Sys2 will normally be running but with very minimal workloads.)
This means that even though Sys2 is physically a replica of Sys1, the initial cost of Sys2 is significantly less than the cost
of Sys1 because the cost of iCAP processors is a fraction of the cost of regularly purchased processors.
At the same time, you also purchase three units of TiCAP for potential use on the failover server Sys2, providing a total
TiCAP balance of 90 processor core days. You apply the TiCAP to Sys2. Figure 4 shows the initial configuration before
a failure.