Adaptive systems permit hardware virtualization and dynamic resource allocation (processors, memory, disk space). Operation and management of applications like SAP is even more efficient on this type of system.

IBM computers employ LPAR (Logical Partitioning) to support this virtualization and allow for parallel operation of different operating systems like AIX 5.1, 5.2, 5.3 or LINUX on a p-Series physical server.

IBM Adaptive Systems are subdivided into several logical servers which appropriate available host resources posing new monitoring challenges. Overall capacities as well as individual capacities and information per partition have to be taken into account.   

Included Workload Managers are only seldom used for dynamic and automated reconfiguration of resources according to application workload. Consequently, the requirements remain after 24/7 monitoring and load analysis of all capacities and, within such systems, now apply to every single partition as well.

This makes it even more important to permanently and comprehensively log workloads per application in order to realize corresponding re-partitioning on the basis of this information.

The following attributes are monitored for the host and per partition and comprehensively analyzed:

• application processes in process groups (virtual storage, CPU time)
• virtual processors (status, idle time, run time)
• memory (virtual and physical storage, size, fill level)
• file systems (size, fill level)
• I-Nodes

• availability and performance – even for large multi-processor UNIX servers
• capacity planning and partition control
• utilization measurements