z/OS und UNIX System Services - Startseite Professur für …st.inf.tu-dresden.de/files/teaching/ss07/ring... · · 2007-06-11z/OS und UNIX System Services Markus Ertl FTSS System

© 2004 IBM Corporation

z/OS und UNIX System Services

Markus ErtlFTSS System z

[email protected]

IBM Software Group

Die wirtschaftliche Bedeutung der zSeries und z/OS

� 95% der weltweit 2000 größten Unternehmen setzen OS/390 f ür

ihren zentralen Server ein

� Zwischen 65% und 70% aller geschäftsrelevanten Daten werd en

auf S/390 Rechnern gespeichert

� 60% aller geschäftsrelevanten Daten, auf die mittels de s WWW

zugegriffen werden kann, sind in Mainframe-Datenbanken

gespeichert

-> nahezu alle deutschen Großunternehmen setzen Rechner der

S/390 und inzwischen der z-Architektur mit OS/390 oder z/OS ein

IBM Software Group

Kundeninvestitionen

� 16.000 Unternehmen weltweit (darunter 490 der Fortune 500

Companies) setzen CICS als Transaktionssystem ein

� 30 Mio CICS Terminals installiert, auf denen täglich 20 Mrd.

Transaktionen ausgeführt werden

� Täglich werden mit CICS Transaktionen 64 Bio Dollar transf eriert

oder abgerechnet

IBM Software Group

Man kann zSeries nicht ohne Betriebssystembetreiben!� z/OS

�The flagship zSeries/z9 operating system

�Ultra-high reliability – over 50% of the OS is dedicated to error recovery!

� z/VM�Run any zSeries/z9 operating system as a “guest” – similar to VMWare, but on a large

scale!

�Sometimes used for application development

� Linux for zSeries & z9�A native Unix operating system for zSeries & z9

�zLinux is already trademarked, so we can’t call it that…

� z/VSE�Small- to mid-size 390 customers

�Most often used on very small 390 hardware

� z/TPF�Primarily for very high transaction rate, specialized applications (airline scheduling)

IBM Software Group

IBM System z9 Virtualization – massive consolidation

JavaAppl.

WebSphere

Business Logic,

Web Serving, ...

Legacy

Database &

Transaction

Processing

Native Linux

DB, Business Logic

Web Serving, ...

ERP

CRM

IBM System z9

MQ DB2IMS CICS

IMS

JavaAppl.Business

Objects

JVM

z/OSz/OS

CICS

DB2

IMS SAPBusiness Objects

Java Appl

Java Appl

Linux

for

zSeries

Applications

CICS

DB2

zVSE

PR/SM up to 60 LPARs100’s to 1000’s !! of virtual linux servers – virtual blades

z/VM

Linux for

zSeries

Linux

Virtual

Servers

HiperSocketsfor integration

Responsive - Variable - Resilient – Focused

z/OS UNIXUSS

Autonomic - Virtualized - Open - Integrated

JVM

Applications

SAP

Oracle

IBM Software Group

S/360 zSeries System zS/390S/370

3081,3083, 3090

3084ES/9000® G4, G5, G6

Assembler C C++

z900 z990 z9 EC

z800 z890 z9 BC

JAVA J2EE

XML

HTML

COBOL

1960s 2000s

OS/360 MFT MVT MVS MVS/XA MVS/ESA

Parallel Sysplex

OS/390 z/OS

1970s 1980s 1990s

MVS Open

Edition UNIX System

Services31-bit24 bit

64-bit MLS

SMS

IMS CICS NCP DB2 WebSphere

IRD zAAP GDPS

HiperSocketsWLM

CUOD HyperSwap

FlashCopy

RTM

Models 40,

50, 65, 90,

91

3145 3158 3031 3032

3155 3168

3033

3195

TSO

OS/VS1

OS/VS2

PR/SM

ISMFISPF

QSAM

BDAMSVS

JES2

JES3

IPO

ServerPac

CBIPO

ACRFRRs

VTAM

DFP

Cross-memory

PDSE

LLA

Hiperspace

RACF

ESCON

FICON

EMIF

z/OS Evolution

Evolution = Application Investment Protection

Americas Advanced Architecture Support

© 2004 IBM Corporation7 Introduction to zSeries – Part 1

z/OS – The ultimate zSeries OS� Components:

– Base control program• Job control for batch – Job Entry Subsystem (JES)• Interactive access – Time Sharing Option (TSO)

– File control and editing via ISPF/PDF

• Workload Manager (WLM) and performance management with RMF

– Security• OS/390 Security Server

– Resource Access Control Facility (RACF)

• Cryptographic services• PKI (Public Key Infrastructure) Services: digital certificates



z/OS Basiskomponenten– Communications/networking

• Communication Server for z/OS• Virtual Telecommunications Access Method (VTAM)

– Enables SNA (Systems Network Architecture) network access to z/OS resources

• TCP/IP for z/OS– Services including FTP, Telnet, NFS, etc.; Hipersocket link

– Storage management• Data Facility Storage Management Subsystem (DFSMS)• Facilities for hierarchical storage management, sort, tape

management, etc.

– Unix System Services• Added in the 1980s (OpenEdition MVS)• Core APIs (Application Programming Interface) in the base operating

system• Hierarchical File System• Interactive Unix shell



TS

O

z/OS Systemstruktur

Base Operating System

Sys

tem

Tas

k

Bat

ch J

ob

TC

P/IP

VT

AM

Bat

ch J

ob

Bat

ch J

ob

Use

rU

ser

Use

rU

ser

IMS

CR

MP

PM

PP

BM

PB

MP

CIC

SA

OR

TO

RA

OR

DO

R

DB

2

Web

Sph

ere

JES

Lotu

s N

otes

LIC (LPAR, etc)

A d d r e s s S p a c e s

� Addressraum Adressierbarkeit – 64-bit in z/OS– 24 bit in MVS/370, 31 bit in MVS/XA –> OS/390

zSeries Hardware



Basics of the Structure of z/OS



Some defining characteristics of z/OS� Uses address spaces to ensure isolation of private areas

� Ensures data integrity, regardless of how large the user population might be.

� Can process a large number of concurrent batch jobs , with automatic workload balancing

� Allows security to be incorporated into application s, resources, and user profiles.

� Allows multiple communications stacks at the same t ime

� Provides extensive recovery, making unplanned syste m restarts very rare.

� Can avoid planned outages for maintenance with Sysp lex

� Can manage mixed workloads

� Can manage large I/O configurations of 10s of 1000s of disk drives, automated tape libraries, large printers, networks of terminals, etc.

� Can be controlled from one or more operator termina ls, or from application programming interfaces (APIs) that allo w automation of routine operator functions.



Large-Scale Processing� Dynamic Workload Balancing

– Within a system image– Across up to 60 system images on one foot print– Across a cluster of up to 32 system images– Automatic adjustment to an increase in batch worklo ad– Transaction goal mode and large process velocity mo de

� Automatic Restart Management (configurable)

– Restarts a failed application on a system image– Restarts a failed application on another system ima ge (within a cluster)

� CICS (TP Monitor) processes more transactions/day than internet webpage hits/day

� Mixed workload capabilities:

– Unix, z/OS Batch, TSO (interactive environment), J2 EE Application Server, TCP/IP stacks, Web Serving …

� Geographically Dispersed Parallel Sysplex for “immediate” Disaster Recovery (up to 100 KM apart)



Super Work Load Management (WLM)� Manages to performance policies and business object ives

– Keeps tracks of who is using what and takes action to get higher priority work done sooner and transactions finished within the IT shops specifications

• Enables classification of work• Enables a persistent priority across all phases of a transaction

– Manages work within a system, across a multi-server cluster across partitions within an LPAR cluster

• Keeps tabs on CPU, Memory and I/O consumption and availability

–– Will swap out work that is in a long wait and may s wap in work tWill swap out work that is in a long wait and may s wap in work t hat needs to finishhat needs to finish

– Will start initiators which can run jobs on the least busy server in the sysplex

– Will route IP traffic to the TCP/IP Stack on the least busy server

– Will give a higher priority to low priority work that is holding a resource needed by higher priority work (ENQ promotion).

– Will create more J2EE servers to handle spikes and meet business objectives

� WLM talks to the hardware (Intelligent Resource Director - IRD)

– Requests reassignment of processors to other partitions

– Requests more I/O channels for band width

– Requests higher I/O priority if I/O requests are not getting assigned



Recovery, Availability and Resiliency � With the first release of MVS in 1976, recovery fun ctions were built in..

– DO NOT BRING DOWN the SYSTEM, IF AT ALL POSSIBLE DO NOT BRING DOWN A COMPONENT, isolate the failure at the lowest possible function.

• Nested recovery routines are supported for applications and system code

• Every key component has a covering recovery routine– Even the recovery routines are covered by recovery routines!!

• Failures are handled, percolated to the next routine, or isolated

– COLLECT DIAGNOSTIC DATA for First Failure Data Capture (FFDC)

• System trace, Component traces and Master (console) trace

• Dumping supported in all environments

• Error symptom strings produced for comparison with previous errors

� Automatic Restart Manager (ARM) to restart critical subsystems and applications



Recovery, Availability and Resiliency � Within a parallel sysplex

– Can recover lost interactive sessions in a heartbeat, you might not even know you lost the connection

– Notification to other systems when a system is “sick” or dead

– Capability to prevent database updates for a “sick” system

– ARM can cause subsystems and applications to be restarted on a surviving server

– Can recover database locks on a surviving system

– Can redirect IP traffic if a TCP/IP stack fails.

– VIPA fall over and VIPA take back is implemented

– Coordinated diagnostic data is gathered on all system in the plex

– And MORE….

• LIKE THE HARDWARE



IMS – Information Management System� Most Corporate Data is Managed by IMS

– Over 95% of Top Fortune 1000 Companies use IMS

– Over 15 million GBs of Production Data managed by IMS

– Nearly $3 Trillion/day transferred through IMS by one customer

– Over 300M savings accounts on-line in IMS at another customer

� Over 50 Billion Transactions a Day run through IMS

– Over 200 Million Users a Day serviced by IMS

– Over 100 Million IMS Trans/Day Handled by One Customer on a single production system

– Over 120M IMS Trans/day, 7M per hour handled by another customer

– Over 6000 Trans/sec across TCP/IP to a single IMS using a single Connect instance

– Over 10 years without an outage at multiple large customers

– Over 21,000 Trans/sec (2 Billion/day) with IMS Data/Queued sharing on a single z990

� Nearly 3M Mips running worldwide with IMS



CICS – Customer Information Control System� 35+ years of applications

� >30B transactions per day

� 5000 packages/2000 ISVs

� 30M CICS users

� 950,000 CICS application programmers

– “it’s the programming model!”

� 90% of IBM’s top 500 customers

� What is it?

– CICS provides an execution environment for concurrent program execution for multiple end users, who have access to multiple data types.

– CICS will manage the operating environment to provide reliability, performance, scalability, security, and integrity



zSeries as a back end WebServerBe prepared for posting highly anticipated files on the WebAlways be able to handle the demand

Pathfinder Mars mission photosDeep Blue - Kasparov matchWoodward trialOlympic games websiteAmerica Online

Secure and scalableWebServer solutionDynamic Workload ManagerReady Support













Nebenbemerkung…wie installiereichdas alles?� No “InstallShield” on z/OS!

� It’s a bit more complex than just running setup.exe

� Software management on z/OS – SMP/E (System Managemen t Program/Extended )

– Industrial-strength installation and inventory– Granular maintenance of system components

• Module-level maintenance• Sophisticated prereq/coreq processing• Installation of APARs (authorized program analysis report)• Tracking of “HOLDDATA” for errors, installation notes, etc.• Batch or ISPF panels for maintenance tasks

– Even applies to Unix System Services and the HFS� “msys for Setup” can be used to simplify initial z/OS configuration

tasks – part of Autonomic Computing initiative



z/OS IBM References �z/OS Home Page – Links to Other Components

� http://www.ibm.com/servers/eserver/zseries/zos/�zFavorites for zSeries ☺☺☺☺

� http://www.ibm.com/servers/eserver/zseries/zos/zfav orites/�z/OS Internet Library

� http://www.ibm.com/servers/eserver/zseries/zos/bkse rv/�zSeries Server Hardware

� http://www.ibm.com/servers/eserver/zseries /�DeveloperWorks – IBM’s Resources for Developers

� http://www-130.ibm.com/developerworks/� IBM AlphaWorks – Emerging Technologies

� http://www.alphaworks.ibm.com/� IBM Technical Journals

� http://www.research.ibm.com/journal/� IBM Redbooks – How-To Books (also Redpieces)

� http://www.redbooks.ibm.com/� IBM Glossary Site� http://www-306.ibm.com/software/globalization/termi nology/index.jsp



z/OS Some Other References� All sorts of MVS Links

� http://www.planetmvs.com� Mainframe Programming, Useful Links for Practioners

� http://www.oberoi-net.com/mainfrme.html� Mainframe Month – Tools and techniques for mainframe professionals

� http://www.xephon.com/mfm/� Mainframe Links

� http://www.mainframelinks.com/� http://www.os390-mvs.freesurf.fr/mainfram.htm

� Mainframe Developer Network� http://www.mainframer.net/

� Mainframes.com Glossary� http://www.mainframes.com/whatis.htm

� Site for System Programmers� http://www.sysprog.net/

� DinosRing� http://n.webring.com/hub?ring=dinos





Questions

Nennen Sie 3 m ögliche Betriebssysteme f ür den Mainframe!

Wieviel Prozessoren und wieviel Hauptspeicher ist maxima l konfigurierbarauf einem modernen Mainframe?

Nennen Sie einen der drei verf ügbaren Spezialprozessoren und seine Aufgabenbereiche!



Documents

z/OS und UNIX System Services - Startseite Professur für …st.inf.tu-dresden.de/files/teaching/ss07/ring... · · 2007-06-11z/OS und UNIX System Services Markus Ertl FTSS System