Can You Teach a New Capacity & Performance Specialist Old Tricks? Hindsight and Insight



Similar documents
Hot Tips from Cheryl & Frank

CA MICS Resource Management r12.7

Cheryl s Hot Flashes #17

z/os Curriculum Job Control Language (JCL) Curriculum JES Curriculum WebSphere Curriculum TSO/ISPF for z/os Curriculum

Exploiting z/os Tales from the MVS Survey

Performance Analytics with TDSz and TCR

Performance Management for

BMC Mainframe Solutions. Optimize the performance, availability and cost of complex z/os environments

z/vm and Linux Disaster Recovery A Customer Experience Lee Stewart Sirius Computer Solutions (DSP)

What s Happening to the Mainframe? Mobile? Social? Cloud? Big Data?

Performance and Capacity Management Best Practices

Big Data Storage in the Cloud

Aktuelles aus z/vm, z/vse, Linux on System z

Berlin Mainframe Summit. Java on z/os IBM Corporation

System z Batch Network Analyzer Tool (zbna) - Because Batch is Back!

CA MICS Resource Management r12.6

zpcr Capacity Sizing Lab Part 2 Hands-on Lab

Predictive Analytics And IT Service Management

Challenges of Capacity Management in Large Mixed Organizations

FAQ: HPA-SQL FOR DB2 MAY

how can I optimize performance of my z/vm and Linux on System z environments?

CICS Transactions Measurement with no Pain

z/os Performance Monitoring Tools Shoot-Out: ASG, BMC, CA, Rocket

Linux System z Why linux experts are choosing the System z platform.

Predictive Analytics And IT Service Management

Consolidated Service Test

z/vm and Linux on zseries Performance Monitoring An Update on How and With What Products

INTERSKILL MAINFRAME TRAINING NEWSLETTER

WebSphere Application Server - Introduction, Monitoring Tools, & Administration

CA SYSVIEW Performance Management r13.0

IBM Infrastructure Suite for z/vm and Linux

Capacity Planning: Where the Mistakes Are Session: 11598

Information Systems. Capacity Planning Monthly Report

z/vm Capacity Planning Overview SHARE 117 Orlando Session 09561

Effective zseries Performance Monitoring Using Resource Measurement Facility

Understanding Linux on z/vm Steal Time

Performance rule violations usually result in increased CPU or I/O, time to fix the mistake, and ultimately, a cost to the business unit.

Computer Associates Unicenter CA-JARS Resource Accounting Software

The mainframe manifesto:

Capacity Management Analytics V2.1

PassGuide.C Questions

IMS Disaster Recovery

Oracle on System z Linux- High Availability Options Session ID 252

Welcome to today's webinar: How to Transform RMF & SMF into Availability Intelligence

Capacity Planning for 1000 virtual servers (What happens when the honey moon is over?) (SHARE SESSION 10334)

CA s Cloud Storage for System z

Effectively running IBM Cognos BI for Linux on z Systems in a z/vm environment

SAP Performance Review/System Health Check

Utility Mainframe System Administration Training Curriculum

Total Cost of Ownership (TCO): Comparing System z and Distributed Platforms

Nalini Elkins' TCP/IP Performance Management, Security, Tuning, and Troubleshooting on z/os

The Comeback of Batch Tuning

Java on z/os. Agenda. Java runtime environments on z/os. Java SDK 5 and 6. Java System Resource Integration. Java Backend Integration

IBM Tivoli Composite Application Manager for WebSphere

IBM AIX MANAGED SERVICES:

Building Effective Dashboard Views Using OMEGAMON and the Tivoli Enterprise Portal

zenterprise The Ideal Platform For Smarter Computing Developing Hybrid Applications For zenterprise

Capacity Planner and Performance Specialist

robert vaupel of mainframe environments

Mainframe. Large Computing Systems. Supercomputer Systems. Mainframe

What s Happening to the Mainframe? Mobile? Social? Cloud? Big Data?

The Total Cost of Ownership (TCO) of migrating to SUSE Linux Enterprise Server for System z

Using Continuous Operations Mode for Proper Backups

Collection of z/os Performance Measurement and Tuning Tips

50 TB Data Warehouse Benchmark on IBM System z

IBM DevOps for Enterprise Systems DevOps for z Systems Is Essential To Deliver Real Business Outcomes

Computer Associates Unicenter NeuMICS Resource Management

CICS Web Services Vendor Shootout

Driving Peak Performance IBM Corporation

IBM RATIONAL PERFORMANCE TESTER

z/os Preventive Maintenance Strategy to Maintain System Availability

Mainframe hardware course: Mainframe s processors

Tips on Monitoring and Managing your z/vm and Linux on System z

P&V Group June The Group P&V is comprised of the brands P&V, VIVIUM, Actel, PNP, Euresa-Life and Arces 1

Mainframe alternative Solution Brief. MFA Sizing Study for a z/os mainframe workload running on a Microsoft and HP Mainframe Alternative (MFA)

Audit & Tune Deliverables

ES&I Operational Team Leader

TITLE: Enhance ESB and BPM solutions with complex data transformation and connectivity for System z

The Journey to Cloud Computing: from experimentation to business reality

Analyzing IBM i Performance Metrics

CA Insight Database Performance Monitor for DB2 for z/os

Buffering, Record Level Sharing, and Performance Basics for VSAM Data Sets

Transcription:

Can You Teach a New Capacity & Performance Specialist Old Tricks? Hindsight and Insight Cheryl Watson Watson & Walker, Inc. March 4, 2009 Session 3117 www.watsonwalker.com home of Cheryl Watson s TUNING Letter, CPU Chart, BoxScore, and GoalTender

Agenda What does a traditional capacity planner do? What does a traditional performance analyst do? What do they have in common? How do they differ? What has changed lately? How to keep learning? 2

What does a traditional capacity planner do? Primary objective plan for changes in capacity (i.e. determine when to get new hardware and how much to get) This is often done only once a year, prior to the budget process Product a Capacity Plan that states the amount of hardware needed, when it s needed, and how the additional resources will be used Hardware processors, memory, storage, links, etc. Additional responsibilities (i.e. how he/she spends the rest of the year!): Interface with CIO to determine his/her needs and timeframes Interface with CEO to understand the company directions and needs Interface with the business departments to determine how business plans might impact IT plans (e.g. new marketing plan goes into effect, big retail sale coming up, large development project on the books) Interface with performance analyst to see if last plan actually worked (i.e. is the last hardware upgrade working as planned?) 3

What does a traditional capacity planner do? Daily Produce & review reports showing resource usage in order to compare estimated usage with actual usage Review performance with performance analyst If discrepancies, determine the reason As Needed Learn about new technologies that might change the next plans (e.g. learn about Linux on System z, learn about ziips & zaaps) Coordinate with business departments to see if their plans have changed Produce reports and information to management when requested Yearly Produce Capacity Plan for management Help performance analyst set service level objectives Avoid vacations prior to budget planning 4

What does a traditional performance analyst do? Primary objective Maintain service levels to meet the objectives Define Service Level Objectives (SLOs) (e.g. 5 minute batch turnaround, 1 second CICS response, etc.) Reports or exception tools can identify when objectives aren t being met Performance analyst determines reason for objectives not being met, and attempts to correct them Additional responsibilities Interface with capacity planner as needed Interface with business departments to set SLOs Interface with operators to identify problems that they are having Manage Workload Manager Policies with sysprog Wear armor and carry a shield as needed 5

What does a traditional performance analyst do? Daily Produce & review reports showing SLOs and key indicators If discrepancies, determine the reason, and resolve Be available to address performance problems as they arise during the day Review performance with capacity planner As Needed Learn about new technologies that might affect performance (e.g. learn about Linux on System z, learn about ziips & zaaps) Produce reports and information to management when requested Learn to use all the monitors Yearly Schedule a vacation when there won t be any performance problems Work with capacity planner to set SLOs 6

What do they have in common? Data SLOs (response & turnaround times) The Capacity Plan is based on a certain level of SLOs; the performance analyst tunes the system to produce agreed upon SLOs Hardware resource usage (volume and response): CPU CPU utilization, CPU total DASD Total terabytes, space by application, CU service times XCF Links activity (peak and average), response times Coupling Facility amount used by application, response time by application, peak and average activity by application Page data sets activity rates, % of usage Application resource usage (volume and response): Number of transactions per day, peak rate of transactions/second Peak and average response times for applications (CICS, IMS, DDF, batch, TSO, MQ, etc.) Amount of CPU usage per application or transaction Business Priorities Interface with charge back staff 7

How do they differ? Time Periods Performance analyst (PA) looks at data by interval (15 30 minutes), by hour, by day, by week PA needs to understand whether key indicators are changing, so he/she also needs to look at key indicators over multiple weeks and months PA primarily looks at response times, but needs volumes to understand the response times Capacity planner (CP) looks at data over longer period of time, such as weekly, monthly, and yearly CP needs to understand if there are daily variations, so he/she can react quickly if something starts to go wrong CP primarily looks at volumes of transactions, CPU usage, DASD usage 8

What has changed lately? Last decade: SMS managed storage Increased DASD caching ziips & zaaps z/vm & IFLs for Linux GDPS IRD & changing number of CPUs (e.g. stopped using average CPU busy, started using total CPU busy) WebSphere applications MQ applications DDF Workload Manager Coupling Facilities Hardware links up the kazoo XCF Storage enhancements zfs Multiple platforms in same reports TCP/IP Automatic schedulers Online monitors RMF/CMF major enhancements PAVs Console management HiperDispatch JES2/JES3 enhancements SDSF enhancements DFSMS products & enhancements DASD hardware enhancements Tape hardware enhancements Lots of new SMF data PDSE VSAM enhancements Request for reports on Internet 9

How to keep learning? Attend SHARE to learn about new and enhanced facilities Attend EWCP sessions at SHARE (Enterprise Wide Capacity & Performance) primarily tuning sessions Look at CMG presentations primarily capacity planning (see www.cmg.org and look at older proceedings) Use IBM Redbooks (search on capacity planning, performance, and tuning) Look for classes on capacity planning and performance PAs really get to understand what your monitors are telling you 10