Changing memory settings
You can view and change the settings of the shared and dedicated memory that is assigned to a partition by using the Hardware Management Console (HMC).
Before you begin
You can change the memory that is allocated to the partition. The views and controls that are presented depend on whether the memory is dedicated or shared, an whether the partition is or running or stopped.
Processors use memory to temporarily hold information. Memory requirements for partitions depend on the partition configuration, I/O resources that are assigned, and applications used.
Memory can be assigned in increments of 16 MB, 32 MB, 64 MB, 128 MB, and 256 MB. The default memory block size varies according to the amount of configurable memory in the system. On systems that are managed by an HMC, memory is assigned to partitions using partition profiles.
Dedicated memory is physical system memory that you allocate to a partition that uses dedicated memory and is reserved for use by the dedicated memory partition until you remove the memory from the dedicated memory partition or delete the dedicated memory partition.
- Number of dedicated memory partitions
- Partition environments of the dedicated memory partitions
- Number of physical and virtual I/O devices that are used by the dedicated memory partitions
- Maximum memory values allocated to the dedicated memory partitions
- Maximum values affect the hardware page table (HPT) size for each dedicated memory partition
- The logical memory map size for each dedicated memory partition
If the server firmware detects that a memory module failed or is about to fail, the server firmware creates a serviceable event. The server firmware can also unconfigure the failing memory module automatically, depending on the type of failure and the deconfiguration policies that you set up by using the Advanced System Management Interface (ASMI). You can also unconfigure a failing memory module manually by using the ASMI. If a memory module failure causes the entire managed system to shut down, the managed system restarts automatically if the managed system is in normal initial program load (IPL) mode. When the managed system restarts itself, or when you restart the managed system manually, the managed system attempts to start the dedicated memory partitions that were running at the time of the memory module failure with their minimum memory values. If the managed system does not have enough memory to start all of the dedicated memory partitions with their minimum memory values, the managed system starts as many dedicated memory partitions as it can with their minimum memory values. After the managed system starts the maximum possible number of dedicated memory partitions, the managed system distributes the leftover memory resources among the running dedicated memory partitions, in proportion to the required memory values of the dedicated memory partitions.
Using Huge pages can improve performance in specific environments that require a high degree of parallelism, such as in the DB2® database. You can specify huge-page memory that can be used for the shared-memory buffer pools in the DB2 database. For logically partitioned systems, you can specify the minimum, wanted, and maximum number of huge pages to allocate to a partition when you create the partition or partition profile.
On managed systems that support huge-page memory, you can use the HMC to set the value for the huge-page memory pool. You can also specify values for the number of huge pages to allocate to partitions.
To use huge-page memory, you must ensure that your system has adequate memory resources to dedicate to the huge-page memory pool. The huge-page memory pool is a region of the system memory that is mapped as 16 GB page segments and is managed separately from the base memory of the system.