User's Manual

Rights Seizure
GiCAP disaster recovery was introduced in Instant Capacity version 8.02.01 and provides the
ability to “seize” core usage rights from a GiCAP member that is off line because of some disaster,
and transfer them to the Group Manager. Then, using normal activation commands, these usage
rights can be used to activate additional processor cores on other group members to increase
capacity. Rights seizure can be used to provide disaster recovery when all partitions of a GiCAP
member fail. It also enables the restoration of usage rights to the member once contact has been
restored.
When a failure occurs on a partition in an active group member, use the icapmanage -x
command to acquire core usage rights from the specified host to make them available to other
group members. This is known as rights seizure. The specified host must be known to the GiCAP
Group Manager (it appears in the output of the icapmanage -s command) and not currently
running. That and any other (vPar) hosts associated with the same hard partition are verified to
be unreachable by the GiCAP Group Manager. The icapmanage -x command verifies with
each known host on the hard partition. The hard partition containing the specified host is left
with one core usage right per active cell. Any core usage rights in excess of this are made available
for use elsewhere in the GiCAP group.
The seizure of core usage rights from a fully unavailable member is temporary. After 10 days,
the usage rights are automatically restored to the member from which they were seized. Seizure
of core usage rights from a member with at least one accessible partition (at the time of rights
seizure) do not expire. The icapmanage -z command allows you to restore previously seized
core usage rights to a specified host.
The following command seizes core usage rights from a partition that is unavailable, so that the
rights are available for other group member activations:
icapmanage -x mypar1.node.corp.com
When to Migrate Usage Rights and When to Seize Usage Rights
Here are some basic guidelines for determining when to seize usage rights or when to simply
migrate those rights:
• Planned downtime and load balancing
Whenever possible, migrate usage rights by deactivating cores in one partition and then
activating cores in another partition. The involved partitions do not need to be part of the
same member server. For example, if maintenance is planned such that a partition is
unavailable for a period of time, deactivate cores in that partition before it becomes
unavailable. The usage rights from this partition are available to the entire GiCAP group
during this maintenance period.
• Unplanned downtime
When an nPartition or server goes down unexpectedly, seize the usage rights from that
nPartition to make them available to the remaining GiCAP group members.
Effects of Rights Seizure
Rights seizure takes almost all the available usage rights from the hard partition containing the
specified host, leaving only enough usage rights to be able to boot (one core usage right for each
cell configured for use-on-next-boot (UONB)). The partition has the value intended active
set to the required minimum (one core per configured cell) and the value actual active set
to zero. The number of core usage rights seized is equal to the difference between the new value
for intended active and the greater of the old values for intended active and actual
active. These seized core usage rights are available for use elsewhere in the GiCAP group.
Although rights can be seized from any hard partition that is unavailable, the Instant Capacity
software makes some additional restrictions when all partitions of a complex are unavailable.
122 Global Instant Capacity