White Paper on Dynamic Processor Deallocation and Dynamic Processor Resilience

NOTE: On N-Class, L-Class and later machines, the monitor will also try to

mark the processor for Deconfiguration whether the processor in

question is a Monarch CPU or not.

NOTE: The EMS CPU monitor will detect and prevent most processor failures

that are related to cache errors. However, although cache errors account

for the majority of processor failures, it is not possible to detect and

prevent all processor related system failures.

For a complete listing on each event that can be generated by the EMS CPU monitor,

refer to the event listings for the CPU monitor provided at:

http://docs.hp.com/hpux/onlinedocs/diag/ems/lpmc_em.htm

Figure 2 – The HP Event Monitoring Service

HP Event Monitoring Service (EMS)

Dynamic Processor Deallocation and Resilience (DPD and DPR) occur in context of

the HP Event Monitoring Service. This is a comprehensive monitoring facility that

Memory

Disks

Tapes

LVM

Datebase MIB

PCX-W

EMS Framework

SNMP

OPC

Syslog

Console

TCP/UDP

HP RC

EMS

Libraries

Notification Methods:

EMS Monitors:

Event Browser

Mail Handler

OpenView IT/O

Syslog File

System Console

Custom Log File

MC/ServiceGuard

Listeners:

Text Log

HP Predictive

Kernel Parms

Events