|
|
|
|
|
|
|
|
Using the HP 3000 Workload Manager: HP 3000 MPE/iX Computer Systems > Chapter 1 Introducing the Workload Manager
Introducing Workgroups |
|
You can think of a workgroup, like its predecessor the scheduling subqueue, as a dynamically changing collection of processes. A workgroup consists of three components: the workgroup name, membership criteria, and specific scheduling characteristics. There are two types of workgroups which you can have on your system:
Each type of workgroup is explained in the following sections. All customers who update to version C.50.00 of MPE/iX will have five system-defined workgroups on their systems. Each system-defined workgroup corresponds to one of the five traditional scheduling subqueues. The AS subqueue, for example, becomes the AS_Default workgroup. Other than the difference in nomenclature, however, there is nothing that distinguishes the system-defined workgroup from their counterpart, the scheduling subqueue. The set of scheduling characteristics that you can control including the base and limit priorities, min and max quantum values, boost property (DECAY or OSCILLATE), and timeslice, as well as their behavior, remains unchanged. Table 1-1 A Comparison of Scheduling Subqueues and System-Defined Workgroups
Each of the system-defined workgroups has one and only one criterion for membership: the queue attribute which specifies its scheduling subqueue. This assures complete backward compatibility with previous versions of the operating system. All user processes will fall into one of these system-defined workgroups using the customary scheduling parameters and all existing programs and job scripts will continue to work. Using the Workload Manager, system managers can create an essentially unlimited number of user-defined workgroups to partition their system's workload. Such workgroups consist of a workgroup name, membership criteria, and scheduling characteristics. Continue reading here for a general introduction to workgroups. For complete information about creating workgroups, modifying workgroups, and changing the workgroup configuration, read Chapter 3. The workgroup name uniquely identifies the workgroup. Typically, it indicates the kinds of processes which become members of the workgroup, such as Program_Development or Payroll_Batch. Membership criteria for a user-defined workgroup can be any combination of the following, but at least one criterion is required:
Each process is assigned to its appropriate workgroup at process creation, whenever a process attribute on which workgroup membership is based is changed, or after it is explicitly assigned to a workgroup. Any command, intrinsic, or AIF that sets or changes the queue attribute of a process may also set or change the workgroup of that process (if QUEUE is specified in the membership criteria.) Subqueue assignment of a process can be set by RUN, HELLO, JOB, and LINK, and it can also be changed with ALTPROC, AIFPROCPUT, and GETPRIORITY. The scheduling characteristics the system manager can assign to a workgroup include the following:
These characteristics, with the exception of minimum and maximum CPU percentage bounds, should be familiar to system managers since they are characteristics of the traditional scheduling subqueues set with the TUNE command. CPU percentage bounds let you either guarantee that processes in a workgroup will get at least a specified amount of CPU time (the minimum CPU percentage) or place an upper limit on the amount of CPU that processes in a workgroup get (the maximum CPU percentage). The next chapter, “Planning for Workgroups”, explains each of these characteristics for individuals who want to review the information conceptually. Chapter 3, “Creating Workgroups”, has examples of how to set and change scheduling characteristics, and all command parameters are fully documented in Chapter 5, “Commands Reference”. |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
||
|
|
|
|||||||||
|
|||||||||
