Resource entitlements

Edit online

WLM allows system administrators to specify per-class resource entitlements independently for each resource type.

These entitlements can be specified by indicating the following:

  • The target for usage of different types of resources. This target is specified with shares. The shares are specified as relative amounts of usage between different classes. For instance, if two classes have respectively 1 and 3 shares of CPU and are the only classes active at this time, their percentage goal used by WLM for its CPU regulation will be 25% and 75%, respectively. The target percentages are calculated for classes in each tier based on the number of active shares in the tier and the amount of resource available to the tier.
  • Minimum and maximum limits. These limits are specified as percentages of the total resource available. WLM supports two kinds of maximum limits:
    • A soft maximum limit indicates the maximum amount of the resource that can be made available when there is contention for the resource. This maximum can be exceeded if there is no contention; that is, if no one else requires the resource.
    • A hard maximum limit indicates the maximum amount of the resource that can be made available regardless of whether there is contention on the resource. Fixed-priority threads, however, are not subject to these same rules and therefore can exceed the limit.
  • Total limits. The total limits are strictly enforced. If a process exceeds one of its total consumption limits, it will be terminated. If a class is at one of its total limits, any operation that would result in the creation of another instance of that resource will fail.

In most cases, soft maximum limits are sufficient to ensure that resource entitlements are met and enforced. Using hard maximum limits may result in unused system resources since these are strictly enforced, even when there is no contention for the resource. Careful consideration must be made when using hard maximum limits since these can greatly affect system or application performance if set too low. Total limits should also be used with caution, since these could result in process termination or failure to function as intended.

In active mode, WLM attempts to keep active classes close to their targets. Since there are few constraints on the values of the various limits, the sum of any of the limits across all classes could far exceed 100%. In this case, if all of the classes are active, the limit cannot be reached by all classes. WLM regulates the processor consumption by adjusting the scheduling priorities of the non-fixed priority threads in the system according to how the class they belong to is performing, relative to its limits and target. This approach guarantees a processor consumption averaged over a given period of time, rather than the processor consumption over very short intervals (for example, 10 ms).

For example, if class A is the only active class, with a processor minimum of 0% and a processor target of 60 shares, then it gets 100% of the processor. If class B, with a processor minimum limit of 0% and a processor target of 40 shares, becomes active, then the class A processor utilization progressively decreases to 60% and the class B processor utilization increases from 0% to 40%. The system stabilizes at 60% and 40% processor utilization, respectively, in a matter of seconds.

This example supposes that there is no memory contention between the classes. Under regular working conditions, the limits you set for processor and memory are interdependent. For example, a class may be unable to reach its target or even its minimum processor allocation if the maximum limit on its memory usage is too low compared to its working set.

To help refine the class definition and class limits for a given set of applications, WLM provides the wlmstat reporting tool, which shows the amount of resource currently being used by each class. A graphical display tool, wlmmon, is also provided for system monitoring.