Chapter 8. z/OS base sysplex on z/VM 101
Global Resource Serialization (GRS) is used to extend ENQ processing to SYSTEMS
(all-members) scope and optionally to convert reserve/release requests that normally
serialize access to an entire volume to ENQ/DEQ requests that serialize specific data set(s).
GRS also uses CTCs to communicate among the members.
8.2.3 What a Parallel Sysplex® is
A Parallel Sysplex looks a lot like a base sysplex, but uses an IBM Coupling Facility (CF) for
signaling among systems. A parallel sysplex extends base sysplex functionality by storing
and retrieving lists of data contained in coupling facility structures. CF structures are used to
implement extended catalog facilities, workload balancing, subsystem functions (for
subsystems like JES2, DB2, CICS, and IMS), and a common system log, among others. CF
can be run only on selected IBM System/390 and z/Series processors.
8.2.4 Why run a base sysplex
Using a Base Sysplex on an EFS system offers advantages in the following areas:
򐂰 Ease of maintenance
򐂰 Isolation
򐂰 Throughput
򐂰 Single-control point
򐂰 Approximation of customer environments
Base sysplex offers the advantages of running more than one z/OS system without requiring
the overhead and work of maintaining separate z/OS installations. Each member of the base
sysplex may be initialized using the same residence volumes and master catalog. Only one
set of target and distribution library data sets needs to be maintained.
The base sysplex offers the ability to isolate workloads from each other; for example, some
programmers might use one image to develop and bench-test code while others use a
second image for integration testing, and a third image runs other work. If one image fails or
is damaged by testing, the other(s) may continue to run.
Global Resource Serialization (GRS) provides systems-scope ENQ/DEQ for members of the
sysplex. GRS may be implemented without defining a base sysplex, but a sysplex cannot be
defined without activating GRS. Under GRS, only the target data set is serialized--remaining
data on a volume may be accessible to other jobs and users. Improvements in throughput are
relative to a shared-spool, shared-DASD configuration that uses reserve/release to protect
shared data. When large numbers of System/390 processors are available (four or more), our
experience suggests that base sysplex may improve overall throughput.
The base sysplex allows an operator to monitor and control all systems from one console.
Commands may be routed to specific systems or broadcast to all systems.
Many EFS systems are used by independent software vendors (ISVs); testing using a base
sysplex may more closely approximate their customers’ environments.
8.3 FLEX-ES system definitions
Our FLEX-ES system definition for running z/VM was:
system sys1:
feature lpar
lparnum(1)
memsize(0x1F0000)

Get S/390 PartnerWorld for Developers, ITSO/EFS Project EFS Systems on a Linux Base: Additional Topics now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.