Chapter 4. Continuous availability and manageability 97
Draft Document for Review September 2, 2008 5:05 pm4405ch04 Continuous availability and manageability.fm
unrecoverable system error, which might otherwise result in an unscheduled server outage.
Dynamic processor deallocation relies upon the service processor’s ability to use
FFDC-generated recoverable error information to notify the POWER Hypervisor when a
processor core reaches its predefined error limit. Then the POWER Hypervisor, in
conjunction with the operating system, redistributes the work to the remaining processor
cores, deallocates the offending processor core, and continues normal operation. Even
reverts from simultaneous multiprocessing to uniprocessor processing.
IBMs logical partitioning strategy allows any processor core to be shared with any partition on
the system, thus enables the following sequential scenarios for processor deallocation and
sparing.
1. An unlicensed Capacity on Demand (CoD) processor core is by default automatically used
for Dynamic Processor Sparing
2. If no CoD processor core is available, the POWER Hypervisor attempts to locate an
un-allocated core somewhere in the system
3. If no spare processor core is available, the POWER Hypervisor attempts to locate a total
of 1.00 spare processing units from a shared processor pool and redistributes the
workload.
4. If the requisite spare capacity is not available, the POWER Hypervisor will determine how
many processing units each partition will need to relinquish to create at least 1.00
processing unit shared processor pool.
5. Once a full core equivalent is attained, the CPU deallocation event occurs.
Figure 4-3 shows a scenario where CoD processor cores are available for dynamic processor
sparing.
Figure 4-3 Dynamic processor deallocation and dynamic processor sparing
To Next Page
Loading...
Need help?
General