Introduction to the new mainframe Chapter 2 Hardware systems and ...

Introduction to the new mainframe Chapter 2 Hardware systems and LPARs

© Copyright py g IBM Corp., p , 2006. All rights g reserved.

Introduction to the new mainframe

Objectives In this chapter you will learn: • • • •

About S/360 and zSeries hardware design About processing units and disk hardware How mainframes differ from PC systems in data encoding About some typical yp hardware configurations g

© Copyright IBM Corp., 2006. All rights reserved.

2


Introduction Here we look at the hardware in a complete system although the emphasis is on the processor “box” T Terminology i l is i nott straightforward t i htf d • Ever since “boxes” became multi-engined, the terms system, processor, p ocesso , a and dC CPU U have a e beco become e muddled udd ed


3


Terminology: System box from IBM possibly a zSeries machine

Individual processors in the system "processors"

Sometimes referenced as a "processor"

"CPUs" "engines"

Sometimes referenced as a "CPU"

A few people use "CEC"

Many use "system"

"PUs" "CPs" IFLs, ICFs, zAAPs, IFLs spares

"system" = CPs running an p g system y operating


4


Conceptual S/360 Storage Control

Processors

Parallel Channels

1

5

3 Control Unit

Channels

A

6

B

3 Control Unit

Control Unit

0

3

1

2

5

Main Storage

3

Devices

0

1

Y

0

1

Z

X C0 Control Unit

communication line

Another System


5


Early system design • System/360 was designed in the early 1960s • The central processor box contains the processors, processors memory, memory control circuits and channel interfaces o Early systems had up to 16 channels whereas modern systems have over 1000

• Channels connect to control units • Control units connect to devices such as disk drives, drives tape drives and communication interfaces


6


Device address address: 1 3 2 channel number

control unit number

device number

In the early y design g the device address was physically p y y related to the hardware architecture Parallel channels had large diameter heavy copper “bus and tag” cables


7


Current design Current CEC designs are considerably more complex then the early S/360 design Thi new complexity This l it includes i l d • I/O connectivity • I/O operation • Partitioning of the system


8


I/O Connectivity CEC box Partition 1

Partition 2

I/O Processing Channels (CHPIDs or PCHIDs)

01 O

LAN

Control unit addresses (CUA)

Unit addresses (UA)

02

...

E

E

42

41

40

E

01 Control Unit

E

...

...

A1

A0

F

F

Other systems

ESCON Director (switch)

C0 Control Unit

0

C1 Control Unit

1

0

FICON switch

01 Control Unit

1

...

...

0

1

02 Control Unit

0

1

E - ESCON channel F - FICON channel O - OSA-Express channel


9


I/O Connectivity • • • • •

ESCON and d FICON channels h l Switches to connect peripheral devices to more than one CEC CHPID addresses are two hex digits Multiple partitions can share CHPIDs I/O subsystem layer exists between the operating system and the CHPIDs


10


I/O Connectivity • I/O controll llayer uses a controll fil file IOCDS that h translates l physical I/O addresses into devices numbers that are used by z/OS • Device numbers are assigned by the system programmer when creating the IODF and IOCDS and are arbitrary (but not random!) • On modern machines they are three or four hex digits


11


I/O Connectivity External device label Four hex digits in range 0000-FFFF Arbitrarily assigned by sysprog Used in JCL, commands, messages, EREP

6830 6831 6832 6833 683F

FF00

HSA LPAR B Central Storage

FF01

LPAR A Central Storage UCB 2001 UCB 2000

FF02

UCB 183F

2000

2008

2001

2009

2002

200A

2003

200B

2004

200C

2005

200D

2006

200E

2007

200F

V 200A,ONLINE

FF03

C40

IEE302I 200A

ONLINE

V 200B,ONLINE


12


System Control and Partitioning Specialized microprocessors for internal control functions

Memory

LPAR1

LPAR2

LPAR3

System Control

HMC

SE

PC

Think Pads

Located in operator area

CP

CP

CP

CP

Processors

System Control

Located inside CEC but can be used byy operators p

Ch Channels l CHPID CHPID

CHPID

CHPID

CHPID

CHPID

CHPID


13


System Control and Partitioning • A Among the h system controll function f i is i the h ability bili to partition i i the h system into logical partitions (LPARs) • Initial limit was 15 LPARs but newer machines allow 30 partitions p • Practical considerations can limit the number to less than this as each LPAR needs memory, I/O, and processing power


14


LPARs System administrators assign: • Memory • Processors • CHPIDs either dedicated or shared

This is done partly in the IOCDS and partly in a system profile on the Support Element (SE) in the CEC. This is normally updated through the HMC. Changing the system profile and IOCDS will usually require a power-on reset (POR) but some changes are dynamic


15


Characteristics of LPARs • LPAR LPARs are the h equivalent i l off a separate mainframe i f for f most practical purposes • Each LPAR runs its own operating p g system y • Devices can be shared across several LPARs • Processors can be dedicated or shared • When shared each LPAR is assigned a number of logical processors (up to the maximum number of physical processors) and a weighting g g • Each LPAR is independent


16


Consolidation of mainframes • M Many iinstallations ll i used d to have h severall boxes b • A single larger machines running many LPARs is often more cost effective,, as software licenses for multiple p small machines can total more than those for a single large one


17


Processing Units Modern M d processor offer ff many specialist i li t processor types. t Most of the additional types are not counted as full processors for software charging • Central Processor (CP) - A full z/OS processor • System Assistance Processor (SAP) - Used for the I/O subsystem – each machine has at least one • Integrated Facility for Linux (IFL) - Special processor for Linux optional • zAAP – Used for JAVA code – optional • zIIP – Used for Dataprocessing code – optional • Integrated Coupling Facility (ICF) - For coupling facilities • Spares

SAPs, IFLs, ICFs, zAAPs and zIIPS are not counted in the model number and or against software costs


18


Capacity on Demand Various forms of Capacity on Demand exist Additional processing power to meet unexpected growth or sudden demand peaks


19


Disk Devices IBM 3390 Disk Unit

channels

IBM 3990 Control Unit

Current mainframes use 3390 disk devices The original configuration was simple with a controller connected to the processor and strings of devices attached to the back end


20


Current 3390 Implementation Host Adapters (2 channel interfaces per adapter) HA

HA

HA

HA

HA

HA

HA

HA

HA

HA

HA

HA

HA

HA

HA

HA

Common Interconnect (across clusters)

Cluster Processor Complex

cache

DA

DA

NVS

DA

DA

RAID array

Cluster Processor Complex

cache

DA

NVS

DA

DA

DA

Device Adapters

RAID array


21


Modern 3390 devices • Th The 2105 E Enterprise i S Storage S Server jjust shown h is i very sophisticated • It emulates a large g number of control units and 3390 disks. It can also be partitioned and connect to UNIX and other systems as SCSI devices. • There are 11 TB of disk space up to 32 channel interfaces, interfaces 16GB cache and 284MB of non-volatile memory


22


Modern 3390 Devices • Th The physical h i l disks di k are commodity di SCSISCSI type units i • Many configurations are possible but usually it is RAID-5 arrays with hot spares p • Almost every part has a fallback or spare and the control units are emulated by 4 RISC processors in two complexes.


23


Modern 3390 Devices • Th The 2105 offers ff Fl FlashCopy, hC E Extended d dR Remote C Copy, C Concurrent Copy, Parallel Access Volumes, Multiple Allegiance • This is a huge g extension of the original g 3390 architecture and offers a massive performance boost. • To the z/OS operating system these disks just appear as traditional 3390 devices so maintaining backward compatibility


24


EBCDIC • Th The IBM S/360 through h h to the h llatest zSeries S i machines hi use the h Extended Binary Coded Decimal Interchange character set for most purposes • This was developed before ASCII and is also an 8 bit character set • z/OS Web Server stores ASCII data as most browsers run on PCs which expect ASCII data • UNICODE is used for JAVA on the latest machines


25


Clustering Clustering has been done for many years in several forms • Basic shared DASD • CTC/GRS rings • Basic and Parallel sysplex

Image is used to describe a single z/OS system system, which might be standalone or an LPAR on a large box


26


Basic shared DASD zSeries ( or LPAR)

zSeries ( or LPAR)

z/OS

z/OS

channels

channels

Real system would have many more control units and devices control unit

control unit

Limited capability y Reserve and release against a whole disk Limits access to that disk for the duration of the update


27


CTC Ring

zSeries ( or LPAR)

zSeries ( or LPAR)

z/OS

z/OS

channels

channels CTC

CTC

control unit

control unit

Can have more systems in the CTC"ring"

Global Resource Reso rce Sharing (GRS) used sed to pass information bet between een systems via the CTC ring Request ENQueue on a data set, update, the DEQueue Loosely coupled system © Copyright IBM Corp., 2006. All rights reserved.

28


Parallel Sysplex

system or LPAR Coupling Facility

CF channels zSeries ( or LPAR)

zSeries ( or LPAR)

z/OS

z/OS

channels

channels

control unit

control unit

This extension of the CTC ring uses a dedicated Coupling Facility to store ENQ data for GRS This is much faster The CF can also be used to share application data such as DB2 tables C appear as a single Can i l system t © Copyright IBM Corp., 2006. All rights reserved.

29


Typical Systems

Standard mainframe control units and devices

Selected mainframe control units and devices

printer i t

printer i t

ESCON channels

Parallel channels FLEX-ES System

MP3000 System

SUPPORT element

LAN p ( ) adapter(s)

tn3270 terminals

LAN adapter(s)

tn3270 terminals

This shows two very small systems • On the left is a Multiprise 3000, which was designed for small installations with internal disk drives • On the right is a FLEX-ES FLEX ES emulation system system, which runs on a PC running Linux or UNIX © Copyright IBM Corp., 2006. All rights reserved.

30


Medium Mainframe Configuration Enterprise Storage Server1

Printer

Printer

z890 OSA-Express Tape drives d i

Router

LAN Enterprise Storage Server1

Consoles

This is a completely modern implementation with no older devices A z890 is featured with two external disk controllers controllers, tape drives drives, printers, LAN attachments and consoles © Copyright IBM Corp., 2006. All rights reserved.

31


Medium Single System This system could be running several LPARs • • • •

A production z/OS with interactive applications A second production LPAR devoted to batch A test z/OS LPAR for testing new software One or more LPARs running g Linux,, perhaps p p running g web applications


32


Moderately Large Configuration Two machines – a new z990 and an older 9672, Parallel Sysplex with coupling facility

CF Consoles

Multiple ESS and older DASD connected via a switch 3745 communications controllers for an SNA network 3490E tape drives retained for compatibility OSA Express connections to various LANs Consoles

OSA Express

LAN

z990

9672 R96

OSA Express OSA Express

OSA Express

LAN LAN

LAN

Escon Director

3490E

Escon Director

Enterprise Storage Server1

Enterprise Storage Server1


Older DASD device 3390

33


Larger system • Thi This is i the h sort off mixture i off device d i generations i that h might i h be b found in any enterprise • New devices are brought g in but older ones are still viable until they reach the end of their life • z/OS will usually run on older devices until an architectural change forces their withdrawal • This sort of change is usually well publicized and several years notice is given of any incompatibility which will usually coincide with the machine being at end of life


34


Summary • T Terminology i l is i important i • The classic S/360 design is important as all later designs have enhanced it. The concepts p are still relevant • New processor types are now available to reduce software costs • EBCDIC character set • Clustering techniques and parallel sysplex


35


FSP Support Element

Bulk Power Unit

FSP

DCA = power + control interface

Bulk Power Unit

Ethernet

Support Elementt El FSP

FSP

FSP

FSP

FSP

FSP

FSP

FSP

DCA

DCA

DCA

DCA

DCA

DCA

DCA

DCA

HMC

System frame "B " "Box"

MCM

MCM

Memory

Memory

Processor Book 0

Processor Book 1

...

Connections for up to 28 channel adapter cards

I/O Cage 2


Connections for up to 28 channel adapter cards

I/O Cage 3

36

Introduction to the new mainframe Chapter 2 Hardware systems and ...

Introduction to the new mainframe Chapter 2 Hardware systems and ...

Suggest Documents

Introduction to the new mainframe Chapter 6: Using Job Control ...

Introduction to the Mainframe - Pearsoncmg

Computer Systems 2: Hardware Interfacing 1. Introduction

Chapter 2 Hardware Accessories

Introduction to the New Mainframe: z/OS Basics

Introduction to the New Mainframe: z/OS Basics - IBM Redbooks

Introduction to the New Mainframe: z/VSE Basics - IBM Redbooks

Introduction to the New Mainframe: z/OS Basics

Introduction to the new mainframe Chapter 10 Topics in z/OS system ...

Chapter 2. Introduction to Statistics and Sampling

Chapter 2 Introduction To Secure Embedded Systems - Springer

CHAPTER 2: COMPUTERS HARDWARE AND SOFTWARE

Ubiquitous Applications: Embedded Systems to Mainframe

Mainframe hardware course: Mainframe's physical ... - Support - IBM

Mainframe hardware course: Mainframe's processors - Support - IBM

Introduction to Hardware Hacking

Chapter 1: Introduction to Foodservice Systems

Chapter 1 Introduction to Fieldbus Systems

Chapter 6 Introduction to Pumping Systems

Chapter 1: Introduction to Operating Systems - Computer ...

Chapter 2: Introduction to Linear Programming

CHAPTER 2 INTRODUCTION TO CLASSICAL PROPOSITIONAL ...

Chapter 1 – Introduction to Operating Systems Chapter 1 ...

SIM900 Mainframe - Stanford Research Systems