Transcript HUES 150: FAMILIES UNDER STRESS

Managing
IT Operations
Dr. Merle P. Martin
MIS Department
CSU Sacramento
Agenda
What is IT Operations?
 IT Operation Trends
 Data Center Operations
 Improving Center Operations
 Data Center Economics
 Outsourcing
 Managing Networks
 Contingency Management

Deloitte & Touche Study
US and Canadian firms
 Late 1995
 431 CIOs
 Abbreviated “DT”

What is IT Operations?
Organization
 Budget emphasis
 Functions

IT Department
CIO
Operations
Systems
Development
Computer
facilities
operations
Data entry
Systems
analysis and
design
Programming
Support
Data
administration
Information
center
Information
technology
CIO Turnover
DT Study, 1995
 1991 14%
 1992 17.5%
 1993 18.8%
 1994 15.5%
 1995 17.0%

Why CIOs Left
Why predecessors left
 Dismissed / demoted 34%
 New position
23%
 Left voluntarily
18%
 Lateral move
10%
 Retired
19%
 Promoted
5%
 44% dismissed from old position

IT Operations Budget
33% Systems and
Programming
 85% maintenance
 15% new development
 decreasing
 10% Administration / Training
 57% IT Operations

Budget by Platform
DT Study, 1995
 Mainframe
44%
 Client / Server 36%
 Other
20%

IT Operations
Functions
Systems operators
 operate hardware
 schedule application runs
 input / output preparation
 Data-entry operators
 Maintenance technicians

IT Operations Trends
Decreasing
hardware costs
 “Lights Out” computer
rooms
 Increasing PCs / workstations
 Expert System (ES) technology
 maintenance troubleshooting

Trends

Increasing use of
automated tools
 Increase in complexity
 Blaab’s Law:
“Old technologies persist in
the face of new technologies”
Improving Data Centers
Application functional
quality
 how it meets user needs
 Technical quality
 how easy to maintain?
 how efficient (MIPS)?
 tension between these goals

Improve Centers

Application age
 legacy systems
 NNC study (1987)
 inefficient centers - 8.7 years
 efficient centers - 4.2 years
Legacy Systems
DT Study, 1995
 65% legacy systems
slated for replacement
were to be replaced over
next 2 years
 Legacy systems still a problem,
but not as significant

Improve Centers

Portfolio coverage
 what % potential
applications are
actually developed?
 beware of “hidden backlog”
 NNC study:
 inefficient centers - 24%
 efficient centers - 40%
Improve Centers
Mix of Center staff:
 NNC study:
 inefficient centers more employees
 greater % of them were
“hands on”
 User feedback
 Problem detection / correction

Issue
17% CIO turnover rate
per year AND 44% were
dismissed or demoted.
 Why would you want
to be a CIO?
 How can you be a
“CIO Survivor”?
What do you think?
Center Economics

“Charge-back” mode
 no operating budget
 charge clients for services
 clients can shop elsewhere
 centers can spend
what they collect - NO MORE!
 Increasing quality
at decreasing cost
Center Economics
Economies of scale
BIG IS BETTER!
 Gorch’s Law
IP = IC ** 2
where:
IP = increase in computer power
IC = increase in computer costs
 Used as barrier to PCs in 1980s

Center Economics

Gorch’s Law works with
Data Centers because:
 Mainframes more efficient
than PCs for large volumes
 3GL more efficient than 4GL
 Telecommunications value =
n (n - 1) where n = # nodes
 Larger network more effective
Center Economics
Typical efficiency measures
 Use Millions of Instructions
per Second (MIPS)
 Hardware Costs per MIPS
 Storage Costs per MIPS
 Software Costs per MIPS
 Teale Data Center “Real Decisions”
study - large IBM

Center Economics

Hardware
 Total Group
$31K / MIPS
 Comparable Size 27K / MIPS
 Govt. Industry
29K / MIPS
 Teale Data Center 12K / MIPS
Center Economics
 Storage
costs - Teale 76%
of Comparable Size group
 Software costs
 Total Group
10.0K /MIPS
 Comparable Size 9.3K /MIPS
 Govt. Industry
11.4K /MIPS
 Teale
8.7K /MIPS
Center Economics

Teale Quality goals:
 > 99.9% availability
 4 hour network Mean
Time to Repair (MTTR)
 Subsecond transaction response
 Immediate Help Desk assistance
Center Economics

Strategies
 Upgrade to latest
equipment / software
 Consolidation
 Outsourcing
Center Consolidation
Chase Manhattan Bank
 100 data centers into 58
 Saved $58 million in 2 years
 First Interstate Bank
 13 data centers into 2
 reduced staff 2300 to 700
 lowered overall budget by
$93 million

Center Consolidation
JC Penny
 20 data centers to 4
 Save $11 million a year
 State of Wisconsin
 12 data centers into one
 Goal: save $1 to $3 million a year
 exempted from state
procurement laws

Outsourcing IT
Functions

WHY?
 overall business trend
 global competition
 value-added criterion
 Would customer pay for this?
 need for technical specialist
 employee costs (fringe benefits)
Outsourcing Types
Professional services
(consulting)
 Services (training / data entry)
 Temporary employees
 contract programmers
 Transactions (credit reports)
 Systems integrators

When to Outsource
Activity not strategic
 Save at least 15%
 Need technology specialists
 Increase financial flexibility
 capital to operating expenses
 free personnel for development
 acquire new technologies quicker

Consolidate,
Transfer
Internally
Improve,
Leverage
“BestofBreed”
Quality
Outsource
Redesign
(Reengineer)
Non - Critical
Critical
Importance
Needs
Improvement
What NOT to
Outsource
Strategy (IT Plan)
 Architecture
 tied to firm’s culture
 Portfolio (What / When)
 Vendor management

Outsource Tendencies
DT Study, 1995
 Disaster recovery
 Network management
 Application development
 Application maintenance
 Data center operations
 Data entry

75%
50%
44%
40%
30%
26%
Outsource Failures
DT Study, 1995
 Vendor expertise and
sophistication
 Cost reduction
 Improved delivery quality
 Increased focus on
core competencies
 Balance sheet improvement

Issue
What IT functions in your
company have been
or should be outsourced?
 Why?
 What obstacles were there
(or are there) to outsourcing?
Network Mgmt Scope
Fault handling
 Performance monitoring
 Change (project) management
 Tactical (future) planning
 Cost control

Network Mgmt Trends
Emerging standards
 Integration as a goal:
 different networks (LAN / WAN)
 different vendors
 Combine with management of
computers
 “finger pointing”

Network (Cont.)
Distributed applications
 client / server
 Oracle’s Lotus Notes
 Increased automation
 Increased outsourcing

Contingency Mgmt
NOT disaster recovery
 reactive, not proactive
 Worst case scenario
 all our eggs in one basket
 natural disaster
 human error / sabotage

Contingency Mgmt.
Methods
Disaster Recovery firm
 outsource strategic function?
 Off-line storage
 Data redundancy
 replicated databases
 fragmented databases

Off-line Storage
Teale Data Center
 Airfreight data twice a
year to Florida (Colorado)
 Send an operator along
 Electronic switching
 Simulate switchover in Oakland
 State of Alaska
 At least 10 miles from site

Contingency Methods
Back-up power generators
 “What if” scenarios
 military war games
 EDS
 Scaled-down manual system
 Back-up / recovery procedures
 Martin, figure 15-12

Record
contents
Transactions
Entry
Backup
records
Online
log files
Recovered
record
contents
End
of
day
Off-line
log file
File
recovery
Contingency Methods
Parallel systems
 Processing backup facility
 Foundation Health
 Contingency Management Plan

How will IT outages affect your firm?
Issue
Tell us a Contingency
Management war story
 What happened?
 How did the firm recover?
 How could the situation have been
 averted?
 mitigated?
Points to Remember
What is “IT Operations”?
 IT Operation trends
 Improving Data Center operations
 Data Center economics
 Outsourcing IT functions
 Managing networks
 Contingency management
These are management issues!
