You are on page 1of 23

Unit objectives

After completing this unit, you should be able to:

• Name some facilities that support high availability of z/OS


systems

• Describe some of the scalability and availability features of


Parallel Sysplex

• Describe some of the data backup and recovery options


provided on z/OS

© Copyright IBM Corporation 2008


Availability and scalability

Availability and scalability

Data backup and recovery

© Copyright IBM Corporation 2008


Maintaining system availability

D
de efin
AC vic e n
TI es ew
VA an
TE d

© Copyright IBM Corporation 2008


IBM Health Checker for z/OS
• Base element in z/OS Base Control Program (BCP):

• Objectives:
– Identify potential problems before they impact availability or, in worst
cases, cause outages
– Check the current active z/OS and sysplex settings and definitions
– Not a diagnostic or monitoring tool
– Continuously running preventive STC that finds potential problems
– Health Checker to produce output in the form of detailed messages
• Potential problems and suggest actions to take

© Copyright IBM Corporation 2008


IBM Health Checker for z/OS architecture
Reports
Product XXX Check
analyzes config for
IBM Health
• Changes in settings or Checker Task
config values
• Threshold levels near Policies HZSPDATA
upper limits Utilities
Data set
• Single points of failure
• "bad" combinations of
config values, settings
Messages

GRS
Check
MONITOR MONITOR MONITOR

RACF Console
XXX GRS RACF
Check h/c log
System Components
© Copyright IBM Corporation 2008
What is a check?
• A check is a program or routine that identifies potential
problems before they impact your system or sysplex
availability:
– Checks are separate from the IBM Health Checker for z/OS
framework.
– Checks contains pre-defined values such as interval, severity, and
routing and descriptor codes.
– Check output is issued in the form of messages. Exceptions produce
write to operator (WTO) messages.
– Check exceptions may be best resolved by running the Health
Checker continuously.
– Checks are managed by printing, activating and deactivating,
refreshing, running, and updating values temporarily or permanently.

© Copyright IBM Corporation 2008


System availability for users
z/OS
Terminal
Regions

Application
Regions

Data Data
CICS
Region Region
Logon

z/OS z/OS
Terminal Terminal
Regions Regions

Application Application
Regions Regions

CICS
Data Data Data Data
Logon Region Region Region Region

© Copyright IBM Corporation 2008


Improved scalability and availability:
Parallel Sysplex

NETWORK

Single system Dynamic workload


image distribution

z/OS z/OS z/OS z/OS

CF CF Continuous
Single point availability
of control

Data sharing

© Copyright IBM Corporation 2008


Parallel Sysplex clustering software structure
DB2
OLTP
(DRDA)
Batch
job

VTAM TCP/IP Single image to network

Transaction managers
Batch Dynamic workload balancing
CICS TSO IMS TM

Appl Appl Appl Appl Appl Appl Applications unchanged

IMS DB DB2 VSAM Oracle Adaplex IDMS Datacom

Data managers Data sharing

z/OS
Base services
Hardware interfaces

© Copyright IBM Corporation 2008


Parallel Sysplex cluster views

Physical view Logical view

Dynamic
Coupling workload balancing
Technology

12
11 1
10 2

9 3

8 4
7 5
6

A A A A Appl
Sysplex timer P P P P
Application
P P P P
L L L L
ESCON

Coupling facility

Shared data
Data
sharing

© Copyright IBM Corporation 2008


Data integrity in a Parallel Sysplex cluster
REQUEST REQUEST
S S

Multisystem
Database manager ySerialization Database manager

yChanged
Locks Data buffers Data buffers Locks
data

MVS Sysplex
z/OS Services z/OS
zSeries zSeries

Coupling technology

yLocks
yDirectories
yCaches

© Copyright IBM Corporation 2008


Coupling facility architected storage

Coupling facility

Cache List Lock

© Copyright IBM Corporation 2008


Database data sharing using coupling facility
Sysplex

DB2 IRLM Coupling facility

SLM Lock
XCF/XES

Cache List
IRLM S X
L C
M F
/
DB2 X
E Database
S

© Copyright IBM Corporation 2008


Workload distribution, routing of transactions

WLM can route transactions to best system, from:


VTAM Generic Resources, Sysplex Distributor for TCP/IP
WLM
VTAM TCP/IP
SNA Generic
Sysplex
Distributor Network
Network Resources

Work
For IMS, CICS, APPC, TSO sessions
Parallel Sysplex
Workload Systems
distribution

Data Coupling Facility

Data sharing

© Copyright IBM Corporation 2008


Data backup and recovery

Availability and scalability

Data backup and recovery

© Copyright IBM Corporation 2008


Maintaining data availability

System Production Test Databases


data data and data and
program program
libraries libraries

© Copyright IBM Corporation 2008


Varying needs of different types of data

System data Test data

yHigh performance yEasy definition


y24x7 availability yAccess security
yReasonable performance
yRecoverability

Production data All data


yAccess security yEnough space
yGood performance
y24x7 availability
yRecoverability
yLong retention
© Copyright IBM Corporation 2008
Life cycle of a data file

Create Reuse

Use
Migrate

Backup

Expire

© Copyright IBM Corporation 2008


Automatic data set or data file availability

DATABASE LARGE PRIME DB2

COMPACTION

BACKUP
BACKUP
© Copyright IBM Corporation 2008
Automatic volume availability

DATABASE LARGE PRIME DB2

DUMP

© Copyright IBM Corporation 2008


Data recovery

VOL1

VOL1’

© Copyright IBM Corporation 2008


Database backup and recovery

DB1

DB1'

Image copy + Logs

© Copyright IBM Corporation 2008


Unit summary
Key points from this unit:

• Dynamic configuration changes and Parallel Sysplex support


high availability of z/OS systems.

• Parallel Sysplex allows systems to be dynamically added to


provide scalability to handle increases in workload.

• Parallel Sysplex provides dynamic workload distribution to


allow work to be routed around systems and applications that
are not currently running.

• Data can be backed up automatically and then recovered in


case of application, system, or media failures.

© Copyright IBM Corporation 2008

You might also like