Planning Volume 2, Control Workstation and Software Environment

Planning for High Availability Cluster Multi-Processing (HACMP)

IBM's tool for building UNIX-based mission-critical computing platforms is the High Availability Cluster Multi-Processing (HACMP) for AIX software package. HACMP ensures that critical resources are available for processing by offering node and application failover and recovery services.

|Four variations of HACMP are available: |

• |HAS - This feature provides the base services for cluster membership, |system management, configuration integrity and control, failover, and |recovery.
• |CRM - This feature optionally adds concurrent shared-access management for |supported RAID and SSA disk subsystems. Concurrent access is provided |at the raw disk level. The application must support some mechanism to |control access to the shared data, such as locking. HAS provides |distributed locking facilities to support this function. The CRM |feature includes the High Availability Subsystem feature.
• |ES - HACMP/ES builds on the Event Management and Group Services components |of PSSP to scale HACMP function up to 32 SP nodes. It also allows |customers to use these facilities to define their own HACMP events. The |Event Management and Group Services facilities provide detection and |notification for loss of a node, network, or adapter. The HACMP/ES |software then drives the appropriate HACMP recovery action. This |feature does not include the High Availability Subsystem, Concurrent Resource |Manager.
• |ESCRM - This feature optionally adds concurrent shared-access management |for supported RAID and SSA disk subsystems. Concurrent access is |provided at the raw disk level. |

Typically, HACMP is run on the control workstation only if HACWS is being used. HACMP can also be run on the SP nodes.

Note:

HACMP does not tolerate IPv6 aliases for IPv4 addresses.

|For complete planning information and for running with HACMP under |various PSSP security configurations, see the book HACMP: Planning |Guide.