Linux and Open Source @ IBM 1 Linux on System z – A Strategic View Len Santalucia CTO & Business...

Linux and Open Source @ IBM

1

Linux on System z – A Strategic View

Len SantaluciaCTO & Business Development Manager

Vicom Infinity, [email protected]

212-799-9375


2

Take back control of your IT infrastructureA data center in a box – not a server farm

Central point of management Increased resource utilization Potentially lower cost of operations

– Less servers– Fewer software licenses– Fewer resources to manage– Less energy, cooling and space

Fewer intrusion points– Tighter security

Fewer points of failure– Greater availability


Virtualization Essentials

x86 systems– Key value proposition: end-user autonomy– “Ctl-Alt-Del” not a problem for a single-user system

UNIX systems– Key value proposition: processor speed – Sweet spot: engineering/scientific computing

Mainframe systems– Key value proposition: mixed workloads– Highest degrees of efficiency, availability, workload mgmt, security

Server Architecture GeneticsConsider the Heritage of Today’s Server Platforms

Virtualization technology can be significantly constrained or compromisedby the underlying system architecture.


Attribute z/VM V5.4 VMware ESX 3.5 System z Value

Supported operating systems Linux, z/OS, z/VSE, z/TPF, z/VM itself Linux, Windows, Netware, Solaris 10 z/VM-on-z/VM = added flexibility

Scalability and Performance

Hypervisor scalability Up to 32 CPUs, 256 GB of memory, More than 1 TB of active virtual memory

Up to 32 CPUs, 256 GB of memory Cost-saving, extreme scalability of virtual server environment

Virtual Machine (VM) scalability Up to 64 CPUs, 1 TB of memory, extensive I/O bandwidth

Up to 4 CPUs, 64 GB of memory, modest I/O bandwidth

Virtualizes servers on z/VM that cannot run on VMware

CPU sharing No limit Up to 8 VMs per CPU Add servers without adding HW

Architected (practical) VM limit Thousands (hundreds) per copy of z/VM

128 (singles) per copy of VMware Avoid real server sprawl

CPU/memory capacity on demand Yes, non-disruptively No Fast, easy capacity growth

In-memory support Minidisk cache; Virtual Disks in Storage; DCSS (shared program executables)

Shared virtual memory pages (detected via background operation)

Enhanced resource utilization

Logical Partition (LPAR) support Yes No Secure Linux access to z/OS

Flexible Operations

Resource over-commitment support (memory, CPU, network, I/O)

Extensive Modest Absorb workload spikes; add more servers to a “full” system

Reconfiguration of Virtual Machines Non-disruptive re-config for CPU, I/O, networking, and memory

VM reboot required for re-config of CPU, memory, ethernet, disk

Higher server and application availability; staff productivity

Command and control, monitoring, automation infrastructure

Extensive, robust, time-tested Modest, yet easy to use Cost-optimized systems management support

Virtual Machine mobility support No; single-image scalability of z/VM does not require mobility for mgmt

Yes; essential for workload mgmt across multiple copies of VMware

Can dynamically add or remove resources to meet demand

Integrity and Security

Fault isolation / hypervisor security Hardware-assisted isolation*;EAL 4+ (CAPP/LSPP)

No I/O virtualization separation;EAL 4+ (No protection profile)

Helps to avoid security breaches; data security and integrity

Run multiple copies of hypervisor on single server

Yes; share CPU, I/O, and networking resources among z/VM systems

No Workload isolation; lower-cost failover (using same hardware)

z/VM runs in System z LPARs, which have achieved EAL 5 certification; System z HiperSockets provide high-speed, secure connectivity among LPARs.*

Functional Comparison of z/VM and VMware ESX


5

The legendary IBM mainframe – IBM System z– Legendary dependability– Extremely security-rich, highly scalable– Designed for multiple diverse workloads executing concurrently– Proven high volume data acquisition and management

The IBM mainframe virtualization capabilities – z/VM 5.3– Improved scalability for applications with large memory requirements– Increased number of virtual guests possible with dedicated devices– … and improved performance and enhanced SCSI disk support

Open standards operating system – Linux for System z– Reliable, stable, security-rich– Available from multiple distributors– Plentiful availability of skills administrators and developers– Large selection of applications middleware and tooling from IBM, ISVs

and Open Source

Linux on IBM System zLinux + Virtualization + System z = SYNERGY


HS

z/VM

A native mainframe operating environment– Exploits IBM System z hardware– Not a unique version of Linux

Application sourcing strategy– The IBM commitment to z/OS, z/VSE and z/TPF is not

affected by this Linux strategy– Customers are offered additional opportunities to leverage

their investments through Linux– New doors are opening for customers

to bring Linux-centric workloads to the platform

What is Linux on System z?

z/OS

HS

z/OS

z/VM


Linux System z Choices

Hipersockets 6GB/S

LPAR LPAR LPAR

z/VM

Linu

x

Linu

x

Linu

x

Linu

x

Linu

x

z/OSz/V

SE

z/V

M

Linu

x

z/O

S

z/O

S

z/VM

LPAR

z10

zAAP CP IFL IFLCPCPCP

zIIP


The Power and Flexibility of System z Virtualization

LPAR 5

z/VM

LPAR 1

z/OS

LPAR 6

Linux

LPAR 8

z/VM

LPAR 2

z/OS

z/OS LinuxLinux

CP zAAP zAAP zIIP zIIP IFL IFLCP CP CP CP CP IFL IFL IFL IFL

LinuxLinuxLinux

Linux ProductionDev/Test and

Optional Failover

LPAR 7

Linux

Performance Criticalz/OS Production

LPAR 3

z/OS

LPAR 4

z/OS

z/OS Pre-Production

IFL IFL

IBM System z Mainframe

Over 40 years of continuous innovation in virtualization technologies Multiple images concurrently share all physical resources Resources delivered as required, automatically, based on business-oriented goals New OS images can be started without affecting ongoing work Hardware assists used to accelerate virtualization operations (e.g., SIE)


9

What System z brings to Linux

The most reliable hardware platform available

– Redundant processors and memory

– Error detection and correction– Remote Support Facility (RSF)

Centralized Linux systems are easier to manage

Designed to support mixed work loads

– Allows consolidation while maintaining one server per application

– Complete work load isolation– High speed inter-server

connectivity

Scalability– System z10 EC scales to

64 application processors– System z9 EC scales to

54 application processors– System z9 BC scales to 7

application processors– eServer zSeries 990

scales to 32 application processors

– Dedicated I/O processors– Hundreds of Linux virtual

servers


10

What is different about Linux on System z?

Access to System z specific hardware– Crypto support – CPA, Crypto2– Traditional and Open I/O subsystems

• Disk (ECKD or SCSI) and tape• SAN Volume Controller

– OSA-Express and OSA-Express3 for very high speed communication between z/OS and Linux

– HiperSockets for ultra-high speed communication between z/OS and Linux on the same machine

z/VM aware– Enhanced performance– System management tools


11

Business Case for Linux on System z

1. Increased solutions through Linux application portfolio

2. Large number of highly skilled programmers familiar with Linux

3. Integrated business solutions Data richness from System z Wide range of Linux applications

4. Industrial strength environment Flexibility and openness of Linux Qualities of service of System z

5. Unique ability to easily consolidate large number of servers


12

Value of Linux on System z

Reduced Total Cost of Ownership (TCO)– Environmental savings – single footprint vs. hundreds of

servers – Consolidation savings – less storage, less servers, less

software licenses, less server management/support Improved service level

– Systems management (single point of control)– Reliability, availability, security of System z

Speed to market– Capacity-on-demand capability on System z– Dynamic allocation of on-line users, less than 10 seconds to

add a new Linux server image using z/VM and IBM DS8000


What z10 EC brings to Linux Customers

4.4 GHz… Quad Core Processor Up to 64 IFLs

Up to 1.5 TB memory

Large Page Support

Hardware Decimal Floating Point

Just in Time Deployment for capacity offerings – permanent and temporary

6.0 GBps HiperSocketsSCSI IPL OSA-Express3 10 GbpsHiperSockets Layer 2 Support


Machine Type – 2098 Model E10– Single frame, air cooled– Non-raised floor option available

Processor Units (PUs)– 12 PUs (3.5GHZ)– 2 SAPs, standard– Zero spares when all PUs characterized– Up to 10 PUs available for characterization

• Central Processors (CPs), Integrated Facility for Linux (IFLs), Internal Coupling Facility (ICFs), System z10 Application Assist Processors (zAAPs), System z10 Integrated Information Processor (zIIP), optional - additional System Assist Processors (SAPs)

Memory– System Minimum of 4 GB– Up to 128 GB for System, including HSA (up to 256 GB, June 30,

2009)• 8 GB Fixed HSA, standard • Up to 120 GB for customer use (up to 248 GB, June 30, 2009) • 4, 8 and 32 GB increments (32 GB increment, June 30, 2009)

I/O– Up to 12 I/O Interconnects per System @ 6 GBps each– 2 Logical Channel Subsystems (LCSSs)– Fiber Quick Connect for ESCON and FICON LX– New OSA-Express3 features

z10 BC Overview


15

IBM Mainframe

CPU 1 CPU 2 CPU 3 CPU 4

Partitioning Firmware

z/VM z/VM z/VM

System z – The ultimate virtualization resource

Massive consolidation platform– 60 logical partitions, 100s to

1000s of virtual servers under z/VM

– Virtualization is built-in, not added-on

– HiperSockets for memory-speed communication

– Most sophisticated and complete hypervisor function available

Intelligent and autonomic management of diverse workloads and system resources based on business policies and workload performance objectives

Utilization often exceeds 90%

– Handles peak workload utilization of 100% without service level degradation


16

z/VM – Unlimited virtualization

z/VM provides a highly flexible test and production environment for enterprises deploying the latest e-business solutions

z/VM helps enterprises meet their growing demands for multi-system server solutions with a broad range of support for operating system environments

Mature technology – z/VM introduced in 1967 Software Hypervisor integrated in hardware

– Sharing of CPU, memory and I/O resources– Virtual network – virtual switches/routers – Virtual I/O (mini-disks, virtual cache, …)– Virtual appliances (SNA/NCP, etc.)

Easy management– Rapid install of new servers – cloning or IBM Director task z/VM

Center– Self-optimizing workload management


17

The value of z/VM for Linux

Enhanced performance, growth and scalability– Server consolidation enables horizontal growth– N-tier architecture on two tiers of hardware– Extensive support for sharing resources – Virtual networking– Effective isolation of Linux images, if required

Increased productivity– Development and testing– Production support

Improved operations– Backup and recovery– Command and control

HS

Linux onSystem z images

Server farms

z/VM


• The majority of virtualization functions are done directly in hardware

• Hardware - saves and loads guests, does address translation, optimizes wait states and spin locks, provides timer facilities, reflects I/O and timer interrupts directly to guests, provides buffer state management for QDIO, allows for second level Hypervisor (z/VM), and other functions

• Results in low latency, low overhead virtualization capabilities

Isolates and dispatches LPARs

Facilitates virtualized networking between LPARs

Uses virtualization hardware

Isolates and dispatches guests

Provides virtualized CPU, memory and I/O for guests

Other guest services

Uses virtualization hardware

z/OS

LPAR

z/VM CP

Linux

Linux

Linux

Linux

Shared or Dedicated CPU LPARs

z/OSz/OS

Shared CPU LPARs

System z Hardware – SIE, Hardware Assists

Hardware Hypervisor (PR/SM)

CMS

z/VM

System z Virtualization

Can run z/VM as a guest under z/VM


Virtualize everything with up to 100% utilization rates– CPU, memory, network, I/O, cryptographic features, coupling facility, ...

Massively scale your workload on a single System z mainframe– The Linux-on-z/VM record is 97,943 virtual machines– Each virtual machine on z/VM can access up to 24,576 devices

Non-disruptively add anything– Up to 64x CPU scalability per mainframe, 32x scalability per z/VM LPAR – z/VM is designed to support more than 1 TB of active virtual memory

Security for everything– Highest security classification for general purpose servers in the world – System z LPAR technology is EAL 5 certified

Optimize and integrate it all with the IBM software portfolio

IBM System z: The Ultimate Virtualization Platform

Consolidate all typesof workloads

Smart economics: start smalland grow big in the same box

Secure your virtualservers and reduce

business risk

Increase staff productivityand virtualize the enterprise

Rapidly respond toworkload spikes


Application serving with Linux on System z

z/VM z/OS

System z

The best LAN is one with

no wires

Internal networkDemilitarized Zone (DMZ)Outside world

Public Key Infrastructure

User

Commerce Server

Caching P

roxy Server

w/ H

TT

P Load B

alancing

Load Balancer w

ith SS

L A

cceleration

Shared File System

Directory Server

Application Node

Collaboration Server

Web Application

Server

Firew

all / LoadBalancer

Systems Management

Database Server

Domain Name Server

Web Application

Server

Intern

et

Firew

all / LoadBalancer

Dom

ain Firew

all

Protocol F

irewall


z/VM V5.4 – An Exceptional Virtualization Platform

z/VM

Linux

Memory

I/O and Network

Linux

LPARResources

CPU

VirtualResources

z/OSCMSLinux

Up to 256 channel paths

Configure virtual machineswith z/VM-unique facilities

z/VM can massively scale a virtual server environment with a mixof virtual and real resources for each virtual machine With exceptional levels of performance, availability, and security Virtual and real assets can be non-disruptively added when needed

Up to 256 GB

Up to 32 CPUs

Optimize virtual servers withdedicated real resources

Add Virtual CPUs (up to 64)

Simulate assets not in LPAR

Up to 24,576 devices pervirtual machine

More than 1 TB of memory(in aggregate)


Users can non-disruptively add memory to a z/VM LPAR– Additional memory can come from: a) unused available memory, b) concurrent memory

upgrade, or c) an LPAR that can release memory– Memory cannot be non-disruptively removed from a z/VM LPAR

z/VM virtualizes this hardware support for guest machines– Currently, only z/OS and z/VM support this capability in a virtual machine environment

Complements ability to dynamically add CPU, I/O, and networking resources

z/VM

Linux

Memory

I/O and Network

Linux

CPU

z/VSE

Smart economics: non-disruptively scale your z/VM environment byadding hardware assets that can be shared with every virtual server

Linux z/VM z/OS

Dynamically addresources toz/VM LPAR

Linux Linux

New with V5.4LPAR

Resources

z/VM Dynamic Memory UpgradeNew z/VM V5.4 Function Enhances System Availability


LPAR

z/VM

CP

z/OS z/OSz/OS

z/VM

LPAR

CP

LinuxLinuxLinux

CPCP

Linux running on z/VM V3

Standard (CP) engines

MLC pricing for z/VM V3

z/OS running on z/VM V3

Standard (CP) engines

MLC pricing for z/VM V3

February 2001February 2001

LPAR

z/VM

IFL

LinuxLinuxLinux

IFL

Linux running on z/VM V4

Integrated Facility for Linux

OTC pricing for z/VM V4

July 2001z/OS z/OSz/OS

z/VM

LPAR

CPCP

z/OS running on z/VM V5.3

Using CPs, zAAPs, zIIPs

OTC pricing for z/VM V5

Sub-cap pricing for z/OSzAAP zIIP

June 2007

September 2008 z/OS z/OSz/OS

z/VM V5.4

LPAR

Linux

CPCP zAAP zIIP IFL

Add Linuxon IFLs onSystem z10

LPAR

z/VM V5.4

IFL

LinuxLinuxLinux

IFL

z/OS

CP

Add z/OSon CPs on

System z10

z/VM and Specialty Engine Support


z/VM-Mode LPAR Support for IBM System z10

LPAR

z/VM

LPAR

z/OS

LPAR

z/VM

LPAR

z/OS

z/OS LinuxLinux

zAAP zAAP zAAP zIIP zIIP ICF ICFCP CP CP CP CP IFL IFL IFL IFL

LinuxLinuxLinux

Linux ProductionDev/Test and Optional Failover

z/OS Production

LPAR

z/OS

LPAR

CFCC

IFL

IBM System z10

z/VM-mode LPAR

New LPAR type for IBM System z10: z/VM-mode– Allows z/VM V5.4 users to configure all CPU types in a z10 LPAR

Offers added flexibility for hosting mainframe workloads– Add IFLs to an existing standard-engine z/VM LPAR to host Linux workloads– Add CPs to an existing IFL z/VM LPAR to host z/OS, z/VSE, or traditional CMS workloads– Add zAAPs and zIIPs to host eligible z/OS specialty-engine processing– Test integrated Linux and z/OS solutions in the same LPAR

No change to software licensing– Software continues to be licensed according to CPU type

CFCC CMSz/OS


Allows z/VM guests to expand or contract the number of virtual processors it uses without affecting the overall CPU capacity it is allowed to consume– Guests can dynamically optimize their multiprogramming capacity based on workload

demand– Starting and stopping virtual CPUs does not affect the total amount of CPU capacity the

guest is authorized to use– Linux CPU hotplug daemon starts and stops virtual CPUs based on Linux Load Average

value Helps enhance the overall efficiency of a Linux-on-z/VM environment

Note: Overall CPU capacity for a guest system can be dynamically adjusted using the SHARE setting

CPU 0SHARE=25

CPU 1SHARE=25

CPU 2SHARE=25

CPU 3SHARE=25

Guest SHARE = 100

CPU 0SHARE=50

CPU 1SHARE=50

CPU 2Stopped

CPU 3Stopped

Guest SHARE = 100

Reduced Need forMultiprogramming

Stop 2 CPUs

CPU 0SHARE=50

CPU 1SHARE=50

CPU 2Stopped

CPU 3Stopped

Guest SHARE = 100

CPU 0SHARE=25

CPU 1SHARE=25

CPU 2SHARE=25

CPU 3SHARE=25

Guest SHARE = 100

Increased Need forMultiprogramming

Start 2 CPUs

Virtual CPU SHARE RedistributionDynamic Virtual Processor Management


Extreme Virtualization with Linux on z/VMLinux Exploitation of z/VM Discontiguous Saved Segments (DCSS)

Linux Linux Linux Linux Linux

VirtualMemory

RealMemory

DCSS support is Data-in-Memory technology– Share a single, real memory location among

multiple virtual machines– Can reduce real memory utilization

Linux exploitation: shared program executables– Program executables are stored in an execute-

in-place file system, then loaded into a DCSS– DCSS memory locations can reside outside

thedefined virtual machine configuration

– Access to file system is at memory speeds;executables are invoked directly out of the filesystem (no data movement required)

– Avoids duplication of virtual memory and datastored on disks

– Helps enhance overall system performanceand scalability

DCSS“A”

DCSS“B”

DCSS“C”

DCSS“A”

DCSS“A”

DCSS“A”

DCSS“C”

DCSS“B”

DCSS“B”

DCSS“B”

Learn more:“Using DCSS/XIP with Oracle 10g on Linux for System z”www.redbooks.ibm.com/redpieces/abstracts/sg247285.html

PGM“A”

PGM“B”

PGM“C”


Extreme Linux-on-z/VM VirtualizationLinux Exploitation of z/VM DCSS Support


VirtualMemory

RealMemory

Discontinguous Saved Segments (DCSS)– Share a single, real memory location among

multiple virtual machines– Can reduce real memory utilization

Linux exploitation: shared program executables– Program executables are stored in an execute-

in-place file system, then loaded into a DCSS– DCSS memory locations can reside outside

thedefined virtual machine configuration

– Access to file system is at memory speeds;executables are invoked directly out of the filesystem (no data movement required)

– Avoids duplication of virtual memory– Helps enhance overall system performance

and scalability z/VM V5.4 support enhancements:

– Segments can reside above 2 GB address line– Enables even greater system scalability– New addressing limit is 512 GB

DCSS“A”

DCSS“B”

DCSS“C”

DCSS“A”

DCSS“A”

DCSS“A”

DCSS“C”

DCSS“B”

DCSS“B”

DCSS“B”

PGM“A”

PGM“B”

PGM“C”

2 GB

Additional DCSS

Addressability

Note: Maximum size of a single DCSS is 2047 MB


System z LPAR

z/VM VSWITCHLACP

Port 1 Port 4Port 2 Port 3

Port 65

z/VM

Port 66 Port 67 Port 68 Port 69 Port 70

Load Balancer Aggregator / Multiplexer

Linux

NIC

Linux

NIC

Linux

NIC

Linux

NIC

Linux

NIC

VMController

OSA OSA OSA OSA

Port 1 Port 4Port 2 Port 3LACP

(Link Aggregation Control Protocol)

Switch

z/VM TCP/IPStack

NIC

z/VM Virtual Switch Link AggregationWith z/VM TCP/IP Stack Connectivity Support in z/VM V5.4

Non-disruptive networkingscalability and failover forGuests and z/VM TCP/IP.

Up to 8 OSA ports per VSWITCH


Start Interpretive Execution- Establish architecture for guest systems- Maintain status- Invoke SIE assists

z/VM – SIE – EAL 4+ – 100s of Virtual Machines – Shared Memory

System z Virtualization TechnologyA Shared Everything Architecture

The potential performance impact of the Linux server farm isisolated from the other LPARs

LPAR Zoning: each partition has a zero-origin address space, allowing I/O access to memory without hypervisor intervention LPAR – Up to 60 Logical Partitions

PR/SM – SIE – EAL 5

Hardware support:10% of circuits areused for virtualization

Most sophisticated and functionally complete hypervisors

Able to host z/OS, Linux, z/VSE, z/TPF, and z/VM-on-z/VM

Shared everything architecture

Highly granular resource sharing (less than 1% utilization)

Any virtual CPU can access any virtual I/O path within the attached logical channel subsystem

z/VM can simulate devices not physically present

Application integration with HiperSockets and VLANs

Intelligent and autonomic workload management

Shared resources per mainframe footprint

Up to 64 OS-configurable CPUs

Up to 10 SAP processors

Up to 1.5 TB of memory

Up to 1024 channel paths

Up to 16 internal HiperSockets networks

HW (LPAR) and SW (z/VM) hypervisors

Hardware support, SIE, microcode assist

Virtualization is transparent for Op Sys execution

Hardware-enforced isolation


IFL Processors

Memory

z/VM

Linux Linux CMS Linux

L P A R

Memory

z/OS

L P A R

Control Program

WebSphere

Memory

L P A R

Processor Resource / System Manager (PR/SM)

Traditional OLTP and

BatchApacheSysAdmin

Tools

WebSphere

Test

HiperSockets & Virtual Networking and Switching

WLM WLM

I/O & Network

Intelligent Resource Director (IRD)

Processors

z/OS

z/VM

z/VSE

OLTP

z/OS

Test

Memory

L P A R

Multi-dimensionalvirtualization technology– System z provides

logical (LPAR) and software (z/VM)partitioning

– PR/SM enableshighly scalablevirtual serverhosting for LPAR and z/VM virtual machineenvironments

– IRD coordinatesallocation of CPU and I/O resources among z/OS and non-z/OS LPARs*

* Excluding non-shared resources like Integrated Facility for Linux processors

IBM System z Virtualization Architecture


31

Workload Isolation – Each user runs in a separate address space– Supervisor state & system programs separated – LPAR separation ensures processing integrity – Storage Protection controls access to protected

areas of storage – HiperSockets communication secures network

communications at memory speed Encryption

– Support for encryption in middleware– Tape Encryption– Key serving– System z cryptographic capabilities

System Integrity Statement– For both z/OS and z/VM– Common Criteria

Scalability– Encryption offload enabled by zIIP – High performance solution

Allows customers to place multiple workloads on single z/OS & Linux Images.

Helps prevent malware, viruses and worms from disrupting systems operations.

Built in Secured System z Processing Reduces Risk


CP

Linux

Console

Linux

Console

Linux

Console

Linux

VirtualConsole

CMS

PROPREXX

MonitorData

HypervisorOperations

CPMonitor

CMS

RealtimeGraphs

Reports,Historical Data

Virtual Servers1. Send all Linux consoleoutput to a single CMSvirtual machine.

1. Use the CP Monitor to automaticallycapture performance and resource consumption data for each Linux server.

2. Use PROP andREXX to interrogateconsole messages.

3. Initiate hypervisorcommands on behalfof Linux servers.

2. Use Performance Toolkitfor VM to process Monitor data.

On-the-flydebug

PerformanceToolkitfor VM

Optimize and Integrate with:- RACF Security Server for z/VM- IBM Director (z/VM Center)- IBM Tivoli OMEGAMON XE for z/VM and Linux- IBM Tivoli Provisioning Manager- IBM WebSphere solutions- IBM Tivoli Monitoring- IBM Operations Manager for z/VM- IBM SAN Volume Controller- More...

z/VM Technology – Command and Control InfrastructureLeveraging the IBM Software Portfolio


IBM Director deployment scope: Templates for z/VM virtual machines and Linux

Provisioning Linux Virtual Machines on System zUsing IBM Director for Linux on System z with z/VM Center


Tivoli Provisioning Manager deployment scope: Operating systems like Linux, AIX, Windows

Middleware like DB2 and WebSphere Application Server

Provisioning Software in System z Virtual Linux ServersUsing IBM Tivoli Provisioning Manager


Combined product offering that monitors z/VM and Linux for System z Provides work spaces that display:

– Overall system health– Workload metrics for

logged-in users– Individual device metrics– LPAR Data

Provides compositeviews of Linux runningon z/VM

Monitoring System z Virtual Linux ServersUsing IBM Tivoli OMEGAMON XE for z/VM and Linux


IBM System z Virtualization Infrastructure

Provisioning M

anagement

Monitoring for Virtualization Infrastructure

Business Services Management

…

Automation for Virtualization Infrastructure

Storage NetworkSecurityExtended Infrastructure Management

Application Layer Management

Resilience M

anagement

IBM Tivoli Virtualization Management for System z Helping Clients Manage and Control Their Virtualized IT Infrastructure


Monitoring for Virtualization Infrastructure• z/VM Virtual Machine Resource Manager (included with z/VM)• IBM z/VM Performance Toolkit for VM (z/VM priced feature)• IBM Director• IBM Tivoli OMEGAMON XE on z/VM and Linux• IBM Tivoli Monitoring• IBM Tivoli Composite Application Manager for SOA• IBM Tivoli Usage and Accounting Manager

Application Layer Management• IBM Tivoli Application Dependency Discovery Manager• IBM Tivoli OMEGAMON XE for Messaging• IBM Tivoli Composite Application Manager for Response Time• IBM Tivoli Composite Application Manager for Web Resources• IBM Tivoli Composite Application Manager for Transactions• IBM Tivoli License Compliance Manager

Automation for Virtualization Infrastructure• IBM Operations Manager for z/VM• IBM Tivoli Enterprise Console • IBM Tivoli Workload Scheduler

IBM System z Virtualization Infrastructure• IBM System z hardware (including LPAR hypervisor)• IBM z/VM Version 5

Business Services Management• IBM Tivoli Business Service Manager• IBM Tivoli Service Request Manager• IBM Change and Configuration Management Database (CCMDB)

Provisioning Management• IBM z/VM DirMaint (z/VM priced feature)• z/VM Center task of IBM Director• IBM Tivoli Provisioning Manager

Extended Infrastructure Management (Security)• IBM z/VM RACF Security Server (z/VM priced feature)• IBM Tivoli zSecure• IBM Tivoli Access Manager for e-business• IBM Tivoli Access Manager for OS• IBM Tivoli Federated Identity Manager• IBM Tivoli Identity Manager• IBM Directory Server• IBM Directory Integrator• IBM Tivoli Risk Manager

Extended Infrastructure Management (Storage)• IBM SAN Volume Controller (SVC)• IBM Tivoli Storage Manager• IBM TotalStorage Productivity Center• IBM Backup and Restore Manager for z/VM• IBM Tape Manager for z/VM• IBM Archive Manager for z/VM

Extended Infrastructure Management (Network)• IBM z/VM RSCS (z/VM priced feature)

Resiliency Management• IBM Tivoli System Automation for Multiplatforms

IBM Tivoli Virtualization Management Portfolio for Linux on z/VM

For specific releases, refer to Tivoli Platform Support Matrix at: ibm.com/software/sysmgmt/products/support/Tivoli_Supported_Platforms.html


38

Customers leveraging scale up and scale out technologies to simplify and integrate their on demand operating environment

As one solution option:– Large SMP and Rack Optimized

servers integrated with Linux, Java and Grid technologies can enable this transformation File/Print

Servers

DNS Servers

DatabaseServers Transaction

Servers

Web Servers

ApplicationServers

Security &Directory Services

File/Print Servers

Scale OutRack Optimized

Scale UpLarge SMP

Application Servers

Collaboration Servers

TerminalServing

SSL Appliances

CorporateInfrastructure

Web Services

E-CommerceApplications

Deep ComputingClusters

JavaLinuxGrid

TransactionData

ReferenceData

Backup Data

SAN

UI Data

DNSServers

Web Servers

ApplicationServers

Security &Directory Servers

File/PrintServers

LAN Servers

DatabaseServers

Business Data

RoutersSwitches

CachingAppliances

SSLAppliances

FirewallServers

Today’s Environment,

Simplified

Infrastructure simplification


39

Ideal blade implementations

Clustered workloads

Distributed computing applications

Infrastructure applications

Small database Processor and

memory intensive workloads

Centralized storage solutions

File/PrintServers


Application Servers


Terminal Serving

SSL Appliances

Infrastructure

Web Services



Virtualization

DNS Servers

DatabaseServers

TransactionServers

Web Servers

ApplicationServers


File/Print Servers

Scale UpLarge SMP


40

DNS Servers

DatabaseServers

TransactionServers

Web Servers

ApplicationServers


File/Print Servers

Scale UpLarge SMP

Virtualization

Ideal mainframe implementations

High performance transaction processing

I/O Intensive workloads

Large database serving

High resiliency and security

Unpredictable and highly variable workload spikes

Low utilization infrastructure applications

Rapid provisioning and re-provisioning

File/PrintServers


Application Servers


Terminal Serving

SSL Appliances

Infrastructure

Web Services




41

Selecting an application

Performance on System z CPUs is superior to any other CPUs on any other platforms– CPU speed is not the entire story – it’s in the architecture!– Architecture designed for multiple or consolidated workloads– System z has definite advantage with applications that have mixed

CPU and I/O– System z10 Enterprise Quad Core 4.4 GHZ

System z and z/VM provide the best virtualization capabilities– All workload types are eligible

Good planning is essential IBM can

– Perform sizing estimate– Perform Application Assessment– Perform zRACE TCO Business Case Analysis – Assist with planning and initial installation needs


42

Technical Considerations Other Considerations

System z “distributed” Application availability– Certification of solution on

hardware/software platform Workload Management Manageability and scaling

characteristics– Especially DB2 on z/OS– Proximity of data to

application– The best network is an

internal network!

Quality of Service

System z “distributed”

Speed of deploymentInstances 2 - n


Data Intensity


Compute Intensity

Where to deploy – System z or “distributed”


Simplified Design Differences Between Distributed and Z

Core Core Core Core

Core Core Core Core

Bus

Healthy CoresAdequate Bus

Great Performance*Good Throughput*

Core

Bus

Core Core Core

Core Core Core Core

Adequate CoresHealthy Bus

Good Performance*Great Throughput*


45

Unify the infrastructure– IT optimization and server consolidation based on virtualization

technology and Linux– Linux can help to simplify systems management with today's

heterogeneous IT environment Leverage the mainframe data serving strengths

– Deploy in less time, accessing core data on z/OS– Reduced networking complexity and improved security network

“inside the box” A secure and flexible business environment

– Linux open standards support for easier application integration– Unparalleled scale up / scale out capabilities– Virtual growth instead of physical expansion on x86 or RISC servers

Leverage strengths across the infrastructure– Superior performance, simplified management, security-rich environment– High-performance security-rich processing with Crypto2 cryptographic co-

processors– Backup and restore processes

Linux on IBM System zTake back control of your IT infrastructure


46

3 year Business Case

0

1

2

3

4

5

6

7

8

Solaris xLinux zLinux

Other

People

SoftwareMaintenance

Software

HW Maintenance

Hardware

•Substantial hardware cost savings, especially with System x

•Software is a major cost driver especially WebLogic and Oracle costs per engine. System x provides no relief

•People costs are significantly reduced Productivity would improve with Linux on System z. System z already has IT process and administration tools to run efficiently

Bottom Line: after considerable “what if” analysis, the conclustion for this situation was Solaris is approximately 3-4 times the 3-year cost of zLinux.Planned IT expense of $1.5 million vs $7.4 million…potential savings 79%


47

$265$640

Adding 1 IFL to a z9 EC 22 x Sun Fire X2100 Opterons

IBM S/W IBM Maint IBM H/W

K$

Potential 3 Year Total Cost of Acquisition (TCA) *

Savings driven by:

Consolidation of OTC SW licenses from 22 to 1

Savings increase as solutions scale up to a 54 way System z

Other likely Savings:

Energy – adding 22 Intel servers consume 6358 watts vs. 0 watts for one incremental IFL*

Space – adding 22 Intel servers requires 5 square feet. Turning on an IFL takes 0 and you can have up to 54 of them (or 789** equivalent Intel servers) with no additional space

People – adding an IFL or multiple IFLs will not likely require more people to manage them

Harness the value of a System z9 EC Mainframe’s high utilization and transform your enterprise’s IT costs

First National Bank of Omaha - Ken Kucera, senior vice president and division head of FNBO Enterprise Technology Services

“For every application I had, I needed another one to five servers behind that, for things like development and application and Web serving. And every 20 servers translates to another body to administer them.”

*Microcode upgrade only** Based on 5% composite Utilization

Leverage an incremental IFL in a Web Serving environment and potentially reduce costs by up to 58% when compared to a equivalent Opteron based Sun Solution

•SW stack includes WebSphere Application Server ND, Linux and zVM•Based on estimated capacity measurements, and 5% server composite utilization for Intel, your actual savings may vary


IT Cost Savings powered by z/VM Virtualization on z10 EC

760 x86 CPs 304 x86 CPs 26 z10 EC IFLs

SpacePowerPeople CostSoftwareMaint

SoftwareMaintHardware

K$

91% Less Hardware 304 x86 Processor Cores vs 26 IFLs

Potential for dramatic reductions in software expense for processor based licenses

Reductions in power and cooling 81% Savings in KWatts and Energy

Costs in this scenario 30% Less Space 93% People savings Increased processor utilization Industry leading Security

Your IT Cost may vary:

Energize your IT savings with z10 EC.

Consolidating 760 Linux serversz/VM versus x86 Virtualization

Oracle DB Workload3-Year Total IT Cost

x86 Virtualization SUN X2100

M2 Dual Core servers

z/VM Linux on System z10

EC

3.8 X

$30.4 M Savings versus

x86 without Virtualization

1.0 X

5.1 X

x86 w/o Virtualization\

SUN X2100 Single Core

servers

80% Savings

Savings 74%


49

When consolidating Linux on low utilization Intel servers, the System z9 Mainframe’s ability to provide high utilization may

help to reduce both power and facility costs

System z10 EC Linux on IntelIn a consolidation,

the System z9 EC

may provide up to

4 times the same

work in the same

space and may

provide up to

12 times the work

for the same

power consumption

The Linux on Intel servers selected in this example are functionally eligible servers considered for consolidation to a System z running at low utilization such that the composite utilization is approximately 5%. The utilization rate assumed for System z10 EC is 90%. This is for illustration only, actual power and space reductions, if any, will vary according to the actual servers selected for consolidation.

KW

an

d S

Q F

ee

t

Processors

0

5

10

15

20

25

30

0 10 20 30 40 50 60

Processors

022 157 319 465 602 789

50

100

150

200

250space

power

space

power

Power and Space Consumption


z/VM Virtualization Value: Power to Simplify

Grow here (inside the box)

Scenario: Host 760 Linux Servers...should I use z/VM virtualization or x86 virtualization?

z/VM Virtualization x86 Virtualization

Grow here (add more boxes!)

One IBM System z10 EC with26 cores (IFLs) and z/VM– with room to add 38 more cores –

x86 blade servers with 304 cores using x86 virtualization product Example: x86 SUN X2100 1U dual-core Opteron 8 racks of 19 dual-core servers per rack running many copies of x86 virtualization product

Simplify your architecture, and simplify management and control.


z/VM Virtualization Value: Environmental Cost

Become Greener with z/VM Virtualization on z10 EC: 5x better than x86

z/VM or x86 Virtualization?

IT Cost Implications of 760-Server

Scenario

z10 EC – 26 IFLs

30 Square Feet

Hourly Energy Usage: 16.3 KWatts

Annual Energy Usage: 0.2M KWatts

Cost: $24.6K/year

z/VM Net Savings per year

900,000 KWatts

$108.4K

81% Less electricity

8 Racks of x86 Blades (304 CPUs)

43 Square Feet

Hourly Energy Usage: 87.8 KWatts*

Annual Energy Usage: 1.1M KWatts*

Cost: $133.0K/year

* Source of power consumption data for the SunFire X2100 (1U) Opteron 2.8 GHz 1 MB server: Competitive Profiles


System Design Affects Virtualization Capabilities Up to 336 I/O Processors

No additional chargefor these processors

System z packs a lot of computepower into a single box With TCO-friendly pricing Up to 64-way SMP

Share up to 64 processorswith up to 60 LPARs

Configure these processorsas CPs, IFLs, zAAPs*, zIIPs*,or ICFs*

* No software license fees

Up to 16 Crypto Express2 CPUsHigh scale performance for SSL transactions

Up to 11 System Assist Processors

Offload system processing to dedicated CPUs (no impact to software license fees)

2 Standard Spare PUs


Why System z Now?

Cost of management & administration

New HW / SW spending

Source: Tony Picardi, IDCEconomist.com: Make it simple. October 28th, 2004From The Economist print edition

15 years ago

Today

20062001

Ops StaffOps Staff

Workload Workload

System z Managing Growth and Complexity

The Linux on Intel servers selected in this example are functionally eligible servers considered for consolidation to a System z running at low utilization such that the composite utilization is approximately 5%. The utilization rate assumed for System z EC is 90%. This is for illustration only actual power and space reductions, if any, will vary according to the actual servers selected for consolidation.

Kilo

wat

ts

Processors

POWER: System z vs. Linux on Intel

Squ

are

Fee

t

Processors

SPACE: System z vs. Linux on Intel

0

50

100

150

200

250

1 vs. 22

8 vs. 157

18 vs. 319

28 vs. 465

38 vs. 602

54 vs. 789

01 vs. 22

8 vs. 157

18 vs. 319

28 vs. 465

38 vs. 602

54 vs. 789

20

40

60

80

100

120

IBM z9

Intel

IBM z9

Intel


zz

z/OSPartitions

IntelIntel

IntelIntel

IntelIntel

IntelIntel

IntelIntel

IntelIntel

IntelIntel

IntelIntel

Traditional discrete servers

App Servers

DB ServerLegacy

HPCCluster

IntelIntel

IntelIntel

IntelIntel

IntelIntel

IntelIntel

IntelIntel

IntelIntel

IntelIntel

Three Ways to Reduce Power Consumption, Footprint & Management Complexity

Consolidated, integrated virtual servers

IFLIFLIFL

Virtual Linux Servers

zz

z/OSPartitions

Hipersockets

IFL IntelIntel

IntelIntel

IntelIntel

IntelIntel

IntelIntel

IntelIntel

IntelIntel

IntelIntel

Reduced Power w/ System z

Virtualization

HPCCluster

Cell BladePool

Cell BladePool

IFLIFLIFL

Virtual Linux Servers

zz

z/OSPartitions

Hipersockets

IFL

Innovative, integrated virtual servers

Reduced Power w/ System z and

Cell

z/OSPartitions

zz IntelIntel

IntelIntel

IntelIntel

IntelIntel

IntelIntel

IntelIntel

IntelIntel

IntelIntel

Traditional discrete servers optimized w/ Cell

App Servers

DB ServerLegacy

HPCCluster

IntelIntel

IntelIntel

Cell Blades

Reduced Power w/

Cell


A B

C

Continuous Availability /

Disaster Recovery Metropolitan Region

Continuous Availability Regionally and

Disaster Recovery Extended Distance

Continuous Availability of Data

within a Data Center

Disaster Recovery atExtended Distance

Near-continuous availability to data

Single Data CenterApplications remain active

GDPS/PPRC HyperSwap Manager

Automated D/R acrosssite or storage failure

No data loss

Two Data CentersSystems remain active

GDPS/ PPRC HyperSwap Manager

GDPS/PPRC

Automated Disaster Recovery

“seconds” of Data Loss

Two Data Centers

GDPS/GMGDPS/XRC

Data availabilityNo data loss

Extended distances

Three Data Centers

GDPS/MGMGDPS/MzGM

The right level of business continuity protection for your business…..GDPS family of offerings


GDPS/PPRC Multiplatform Resiliency for System z

Lin1

z/VM

LPAR1

Lin2 Lin3 Lin4

IBM Tivoli System Automation (SA) for Multiplatforms

CICS& DB2

z/OS

LPAR2

SA z/OSNetView

SAP DBServer

z/OS

LPAR3

SA z/OSNetView

GDPS System

z/OS

LPAR4

SA z/OSNetView

LPAR5

SAP App Servers

Site 1 Site 2

PPRC

Site

z/OS Sysplex

Expend-able

Workload

Designed for customerswith distributedapplications

SAP application serverrunning on Linux forSystem z

SAP DB server runningon z/OS

Coordinated near-continuous availabilityand DR solution forz/OS, Linux guests, and z/VM

Uses z/VM HyperSwapfunction to switch tosecondary disks

Sysplex support allowsfor site recovery

Takeover


z9-EC – Under the covers

HybridCooling

Processor Books and Memory

CEC Cage

STI cables

SupportElements

3x I/Ocages

PowerSupplies

InternalBatteries

Front View

Fiber Quick Connect Feature


InternalBatteries(optional)

PowerSupplies

3x I/Ocages

Fiber Quick Connect (FQC) Feature

(optional)

Processor Books, Memory, MBA and

HCA cards

2 x CoolingUnits

InfiniBand I/O Interconnects

2 x SupportElements

FICON & ESCON FQC

Ethernet cables for internal System LAN connecting Flexible Service Processor

(FSP) cage controller cards

z10 EC – Under the covers (Model E56 or E64)


z10 BC Frame layout in detail


InternalBattery

(optional)

PowerSupplies

4 x I/O Drawers

Fiber Quick Connect (FQC) Feature

(optional – not shown)

CPC (SCMs, Memory, MBA, HCA and FSP )

Drawer2 x SupportElements

2x Support Elements

FICON LX & ESCON FQC

z10 BC – Under the covers Front View

I/O Drawer #3

I/O Drawer #2

I/O Drawer #1

I/O Drawer #4


Single PU Chip

without heatsink

MCM– 96mm x 96mm in size– 5 PU chips per MCM

• Quad core chips with 3 or 4 active cores• PU Chip size 21.97 mm x 21.17 mm

– 2 SC chips per MCM• 24 MB L2 cache per chip • SC Chip size 21.11 mm x 21.71 mm

– Up to 4 MCMs for System

PU 0PU 2

PU 4 PU 3

SC 0SC 1

PU 1

S 0

S 1

S 2

S 3

z10 EC MCM PU SCM

– 50mm x 50mm in size – fully assembled– Quad core chip with 3 active cores– 4 PU SCMs per System with total of 12 cores– PU Chip size 21.97 mm x 21.17 mm

SC SCM– 61mm x 61mm in size – fully assembled– 2 SC SCMs per System– 24 MB L2 cache per chip – SC Chip size 21.11 mm x 21.71 mm

z10 BC SCMs

Single SC Chip

without heatsink

z10 BC SCM Vs z10 EC MCM Comparison


MSC MSCPU PU

PU

PU

PU PU

PU

PUSD SD

SD SD

SCCLK

System z9 EC/BC Multi Chip Module (MCM)

Advanced 95mm x 95mm MCM– 104 Glass Ceramic layers– 16 chip sites, 217 capacitors– 0.476 km of internal wire

CMOS 10K chip Technology

– PU, SC, SD and MSC chips– Copper interconnections, 10 copper layers– 8 PU chips/MCM

• 15.78 mm x 11.84 mm, 121 million transistors/chip• L1 cache/PU

– 256 KB I-cache, 256 KB D-cache• 0.58 ns Cycle Time

– 4 System Data (SD) cache chips/MCM• 15.66 mm x 15.40mm• L2 cache per Book

– 660 million transistors/chip, 40 MB– 1 Storage Control (SC) chip

• 16.41mm x 16.41mm, 162 million transistors• L2 cache crosspoint switch• L2 access rings to/from other MCMs

– 2 Memory Storage Control (MSC) chips• 14.31 mm x 14.31 mm, 24 million transistors/chip• Memory cards (L3) interface to L2• L2 access to/from MBAs (off MCM)

– 1 Clock (CLK) chip - CMOS 8S• Clock and ETR Receiver

4 Billion Transistors

Coolest and Densest Logic Package


96mm x 96mm MCM–103 Glass Ceramic layers–7 chip sites–7356 LGA connections–17 and 20 way MCMs

CMOS 11s chip Technology – PU, SC, S chips, 65 nm– 5 PU chips/MCM – Each up to 4 cores

• One memory control (MC) per PU chip• 21.97 mm x 21.17 mm• 994 million transistors/PU chip• L1 cache/PU core

– 64 KB I-cache– 128 KB D-cache

• L1.5 cache/PU core– 3 MB

• 4.4 GHz• 0.23 ns Cycle Time• 6 km of wire

– 2 Storage Control (SC) chip• 21.11 mm x 21.71 mm• 1.6 billion transistors/chip• L2 Cache 24 MB per SC chip (48 MB/Book)• L2 access to/from other MCMs• 3 km of wire

– 4 SEEPROM (S) chips• 2 x active and 2 x redundant• Product data for MCM, chips and other engineering

information– Clock Functions – distributed across PU and SC chips

• Master Time-of-Day (TOD) and 9037 (ETR) functions are on the SC

z10 EC Multi-Chip Module (MCM)

PU 0PU 2

PU 4 PU 3

SC 0SC 1

PU 1

S 0

S 1

S 2

S 3


66 Apr 18, 2023

ProtecTIER software resides on TS7650G Deduplication Gateway Emulates a tape library unit, including drives, cartridges and robotics Uses FC-attached disk array as the backup medium

TS7650G ProtecTIER Overview

Backup Server

FC

Virtual Tape Library

TS7650G

Disk Array


Repository

Backup Servers

FC Switch TS7650G

HyperFactor

MemoryResident Index

“Filtered” data

Existing Data

New Data Stream Ratio of Repository to Index: 250,000 : 1

The Impact of HyperFactor

Disk Arrays


System z – PU Characterization

Processor Characterizations– Define processors for unique requirements– All processors are the same hardware– Specialty processors operate at full speed– CPs have sub-capacity options

Benefits– Lower software and hardware costs– Satisfy unique requirements– Spares are always present for availability

zAAP zIIP SAP ANY

IBM System z

CP

CentralProcessors

z/OSz/VMLinuxTPF

z/VSE

LPARs

CP zAAP

zSeriesApplication

AssistProcessor

JAVAOffload

zIIP

Z9 Integrated

InformationProcessor

DatabaseIPSec

Offload

SAP

SystemAssist

Processor

Part ofChannel

Subsystem

ANY

SPARES

CUoDGrowth

andRAS

ICFICF

Integrated CouplingFacilityParallel SysplexCouplingFacilityCode

LPARs

IFLIFL

IntegratedFacility for

Linux

Linux orz/VM

LPARs Offload Offload Offload


Linux as a share of System z

– There are over 1300 System z customers with Linux MIPs (install base)

– Linux is currently 10% of hardware revenue (as of year end ‘07)

– Linux is currently 18% of MIPs (installed as of year end ‘07))

– Linux has penetrated 17% of System z installed customers (16% year end 2006)

– 59% of our top 100 mainframe customers have Linux ( 45% year end 2006)

– The top 100 Linux customers average 16 IFL’s (9 year end 2006)

• The top 100 Linux customers average w/o 3

large Japanese account average 11 IFL’s

Success and Targets With Linux on System z

1999 2000 2001 2002 2003 2004 2005 2006 2007 20080

1000

2000

3000

4000

5000

6000

7000

8000

Accumulated accounted IFL Shipments

Projected target growth rates for 2008 off of 2007 Actuals

+14% number of IFL engines (1262)+10% Linux attributed revenue +49% Linux capacity


70

Customer Patterns1. Consolidation of simple web-, application-, file-, print-serving

– Customer objective: “try it out” very limited z/OS backend integration very small footprint (1 IFL only), no real mission-critical workload deployed

– All kind of customers types (very small to very large, all sectors)

2. Migration of costly distributed Application Server infrastructure for z/OS DB2-backend processing– Customer objective: “reduce my TCO & get better controls”

tight z/OS integration easy to achieve technical benefits (superior transaction rates & RAS

characteristics) Immediate Systems Managements benefits thru central point of administration Examples: SAP, WAS, WCS, WPS, S2, BEA WebLogic, IBI WebFocus, …

– Primarily FSS and large IND, DIS, COMS customers

3. Infrastructure Simplification– Customer objective: “get back on track in handling distributed environment”

indifferent z/OS integration Customers cannot manage constant growth of distributed infrastructure in terms

of staff, skills, environmentals, controls– Primarily SMB, PUB and small other sector customers

4. Migration of mission-critical end-to-end applications– Customer objective: “run it on the most reliable and most secure platform”

no or very limited z/OS integration Current hosting infrastructure for mission critical distributed Apps too unreliable or

insecure Scope is on Multi-Tier workload (App Servers + DB Server + Front End Servers +

Applications), currently hosted on Unix or Windows platforms Superior RAS, BR & Security characteristics

– Primarily very large FSS, IND, DIS, COMS customers

Strategy

68%

21%

11% 0%

10%

38%

28%

24%


71

Nationwide

Key Benefits (Value Proposition)Expect to save over $15M over the next 3 yearsSavings will be in cooling, maintenance, software and equipment costs, said Guru Vasudeva, a

Nationwide computer expert who is overseeing the technology's implementation. Lower middleware and application costs, 50% reduction in monthly charges for Web infrastructure

80% reduction in data center floor space utilization, optimized CPU utilizationGreater operational and managerial efficiencies and lower cost per virtual server Building better capacity management processes and workload modeling to better assess which

applications and workloads most appropriate to migrate to the z platform for additional cost savingsLeveraged IBM services, server and software expertise for best practices in tuning and capacity

management, better management and resource optimization to drive down costs

Solution► GTS Capacity Planning and Capacity Management Services► IBM z9 IFLs and associated systems software licenses► Novell SUSE Enterprise Linux ► IBM WebSphere► IBM DB2 Universal Database™ (UDB)► IBM WebSphere MQ► SupportLine Linux support


72

First National Bank of Omaha

Large complex IT infrastructure was difficult to monitor, manage and scale

With 600 servers, maintenance costs skyrocketed

Staff growth 30% each year Average 12% server and 14% storage

utilization rates Peak transaction volumes in certain

applications forced bank to continually add capacity

IBM System z with z/VM and Linux virtual servers70 IBM BladeCenterTM serversIBM SAN Volume ControllerIBM System p 695 runs the bank’s data warehouse

Challenge

Solution

Benefits:• Realized ratio of 18 : 1 on physical server

consolidation by using virtualization

•Ported 30 SUN based WebSphere server applications to Linux on System z

•Consolidated 560 Intel-based servers to 70 BladeCenter servers

• Decreased systems staff from 30 to 8 to manage entire infrastructure

• Capacity Upgrade On demand to provide additional mainframe computing resources when extra capacity required

• Storage consolidation helped bank reduce number of devices to manage

“As a result of consolidation, we only have to maintain a handful of servers instead of nearly 600 – making the task much less complex and expensive” Kenneth J. Kucera, senior vice president, division head of Enterprise Technology Services


73

Key Benefits (Value Proposition) An architecture that suits requirements for security, manageability,

reliability, availability, scalability, extensibility and flexibility The ability to grow Nexxar ‘s growth by acquisition business while staying

within the same platform Consolidation of more than 80 x86 servers onto an IBM System z9 Business

Class (BC)A 75% reduction of headcount required to maintain the operating

environment in comparison with the x86 systems previously on the floor.

Business NeedA architecture for IT infrastructure to provide very high (24x7) availability and the ability to sustain significant anticipated business growth

Nexxar

Solution Hardware IBM System z9

– Storage (DS8100, 3590)

Software zOS-DB2, zVM-Linux– WebSphere Application Server– Tivoli OMEGAMON– Rational

Services GTS Infrastructure & Systems Management Services

Advance virtualization capabilities to quickly create a secure, custom-tailored computing environment for each “private label” relationship


74

Customer issues ：

IBM solution:

IBM System z9™ 109

Advantage:

•A lot of Unix and x86 servers increase TCO. •A lot of CPUs increase Oracle DB license cost.•Physical distributed DB ( Oracle ) decreases Performance.•Dividing DB in Unix & x86 servers difficult to add new application.

•Server consolidation solution with IBM System z9 and Novell SLES for the mission critical online stock order system. •Virtualization technology of IBM System z9 enables many Unix & x86 servers to be consolidated into multiple IBM System z9’s( 54 IFLs). Providing the environment by Linux where multi-vendor SW runs.•Expanding the DB information system with z990 ( 32 IFLs ). •IBM Linux support by IBM Linux Technology Center and Global team.

Reduce cost of SW license Providing Non-stop 24 x 365 system of online stock order system Providing the environment for Oracle 10g RAC High transactions ( more than 1,000 Trx /s ) with IBM system z9 advantages. The Linux support level is stellar.

IBM Linux Solutions

Large Japanese Financial Firms and Banks

Very Large DB consolidation for Online Stock order and Banking Systems


IBM consolidates its own data centers for large savings

IBM Global Account (IGA) IT CostsVaried Distributed Workloads5-Year IT Cost Study Results

3,900 x86 & UNIX servers

30 z10 ECIFL servers

Potential 5-Year IT Cost savings

5-Y

ea

r IT

Ex

pe

ns

es

(K

$)

BEFORELinux on System

z10

Potential 5 Year Total IT Cost*

Workload consolidation using Linux on a

mainframe may result in over 40% IT Cost savings

(your IT costs may vary)

92% less hardware

23,000 processor cores going to 1,782 IFLs

+80% energy reduction

+85% space reduction

180% increase in utilization

Reduced People cost through virtualization

Dramatic reductions in software expense

Significant reductions in IT Data Center square footage

Enables growth

Better utilization of facilities


76

Financial / InsuranceBank of Tokyo MitsubishiFed ReserveFirst Nat Bank OmahaDeutsche BankE*TradeHewitt AssocINGWells FargoWachoviaHSBC AXAAMEXState Street BankJPMorganChaseLehmanFiServMorgan StanleyFreddie MacADPCitizens BankNorthern TrustBrown Brothers HarrimanCitigroupGoldman SachsMerrill LynchFidelity InvestmentsBank of MontrealNYFIXNYSE/MYMEUBS WarburgCheckFreeNomuraCharles SchwabCredit SuissePrudential SecuritiesBank of AmericaBankOneBanco de BrazilEmigrant BankNexxarVISATransunion

ManufacturingElektro-Material AGDaimlerChryslerFERMAX Electronica, S.A.E.Ford Grede Foundries, Inc.Hero-Honda Motors, LtdKodakMotorolaNewell RubbermaidCanonSamsungVolvoWinnebagoYamaha CanadaPanasonicGEHoneywellSharpDistribution/RetailFedEx UPSBaldorKB ToysKonica Business TechnologySafewayMacy'sSUPERVALUPublixBoscov'sBurlington Coat FactoryHannaford BrosBarnes & NobleCendantCircuit CityL.L. BeanPrice ChopperSherwin-WilliamsWinn DixieArrow ElectronicsPep BoysAramark

EducationCollege of New JerseyArizona State UniversityBinghamton UBoston UniversityBuffalo UClarkson UU ConnClemsonColorado State UniversityNYUThomas Jefferson UMarist CollegePenn State UPalm Beach CCUniversity of NebraskaUniversity of GenevaUniversity of New MexicoUniversity of FloridaCUNY

PublicNYPD, NYCHA, NYC DoITT LA St Louis Orem Richmond Philadelphia SacramentoMiami Dade CountyPinellas CountyGermany China Braz UK Italy USA

Aus Japan Can Ind Korea QuebecOnondaga CountyPA DE VT NC MA OK WI NJ TX HI NY VA TN

ServiceAkamaiASPCALexis NexisMarriottVeriSignEXXON Mobile Travel Guide

MediaNY TimesHoplonWeather.com

ISPs/ASPsAleosASP MoversBRFreeCOMSAT VenezuelaCLS ARGOSDimension EnterprisesDreamballExxcomFileFlowiMobile Computing, Inc.Infocom Technologies, Inc.Infocrossing/ACTSIntelnetInterServJindal Information Sys.Mazu NetworksNarmada Commun.Oni WayQnet InfosystemSERCONSonera EntrumWorldGate Private Networks

Teleco Swisscom

Verizon Wireless, VerizonATTBell CanadaTelemar

Federal USPS, EPA

DOE, DODUSDAPentagon, Homeland SecuritySecret Agencies/Projects

TransportationAir New ZealandDelta Air LinesCendantExxon Mobil Travel GuideKorean AirlinesThrifty CarNorfolk Southern RR

PharmaceuticalBristol MyersAventisAstra ZenecaSchering-PloughMerckGlaxoSK

TechnologyBrookhaven National LabChevronGene FormaticsJet Propulsion LaboratoryPhillips PetroleumRoyal Dutch ShelllStructural BioinformaticsComputer AssocBMC

HealthHMAMt Sinai HospSloan Kettering HospU of R HospitalSUNY Med Ctr SyracuseUCLA HealthcareCornell U Med CtrSiemens

InsuranceAmerican Fidelity AssuranceNationwide InsNCCIGuideOne InsUnivarMet LifeLiberty MutualAFLACMass MutualGeico InsJackson National Life InsBCBS TN/FL/MN/MD/Highmark/KCUtah InsAIBFarmers InsArbella InsZurich/Farmers InsWisconsin Physicians SrvcsFirst Ins of Hawaii


77

Next Steps?

Executive Sponsorship Linux on IBM System z Pilot Project Plan w/Success Criteria Consider Linux on System z to

lower costs, increase reliability and security, address compliance, provide superior disaster recovery and business continuance, and improve service


Backup Material


Extreme Virtualization with Linux on z/VMVMRM Cooperative Memory Management (VMRM-CMM)


= Active virtual memory

= Inactive virtual memory

VirtualMemory

RealMemory

ExpandedStorage

Disk Space

z/VM PagingSubsystem

Problem scenario: virtual memory utilization far exceeds real memory availability

Solution: real memory constraint corrected by z/VM Virtual Machine Resource Manager

Linux images signaled to reducevirtual memory consumption

Demand on real memory andz/VM paging subsystemis reduced

Helps improve overall systemperformance and guest imagethroughput

Virt

ua

l Ma

chin

eR

eso

urc

e M

an

ag

er

Learn more at:ibm.com/servers/eserver/zseries/zvm/sysman/vmrm/vmrmcmm.html


OLTP Database Environment with VMRM-CMM and CMMAExcerpt from “z/VM Large Memory – Linux on System z” Whitepaper

50% MoreThroughput


Linux and z/VM Technology ExploitationCollaborative Memory Management Assist (CMMA)


= Stable

= Unused

VirtualMemory

RealMemory

ExpandedStorage

Disk Space

z/VM PagingSubsystem

= Volatile discarded

= Volatile

Extends coordination of memory and paging between Linux and z/VM to the level of individual pages usinga new hardware assist (CMMA)

z/VM knows when a Linux application hasreleased a page of memory

Host Page-Management Assist (HPMA), in conjunctionwith CMMA, further reducesz/VM processing needed toresolve page faults

Can help z/VM host morevirtual servers in the sameamount of memory

Supported by System z9and z/VM V5.3

Linux support availablewith Novell SLES 10 SP1


Extreme Virtualization with Linux on z/VMLinux Exploitation of z/VM Virtual Disks in Storage (VDISK)

VirtualMemory

RealMemory

VDISK support is Data-in-Memory technology– Simulate a disk device using real memory– Achieve memory speeds on disk I/O operations– VDISKs can be shared among virtual machines

Linux exploitation: high-speed swap device– Use VDISKs for Linux swap devices instead of

real disk volumes– Reduces demand on I/O subsystem– Helps reduce the performance penalty normally

associated with swapping operations– An excellent configuration tool that helps clients

minimize the memory footprint required for virtual Linux servers

– Helps improve the efficiency of sharing real resources among virtual machines


VDISK VDISK VDISK VDISK VDISK

Linux and Open Source @ IBM 1 Linux on System z – A Strategic View Len Santalucia CTO & Business...

Documents

Transcript of Linux and Open Source @ IBM 1 Linux on System z – A Strategic View Len Santalucia CTO & Business...