System design and implementation information technology question

profilejackr.0035
Lecture17-GoLive.pptx

Lecture 17 – cutover and go live

1

schedule

Before you “go live” a lot of planning for that date has to have taken place

Create schedule

The schedule should include:

Of course, setting the date

Notifying all affected personnel not involved in the cutover team – depending on the system, this could be the entire company – or not

Completing all system documentation

Training users on test system – when and where this will be done

Backing up all programs that are being modified – sometimes this means the entire business system

Backing up all data that will be involved – unlike backing up the programs, this may have to be done as close to cutover as possible

Backing up the data as close as possible to cutover date (or dates) may not always be possible.

Ensure you have correct data

What if your data is backed up BEFORE a system transforms it, and then you perform last processing, based on a time (daily, weekly, or monthly).

Then you install system, and load THAT data. What happens?

2

Four installation strategies

The strategies

Direct cutover or installation

Parallel operation or installation

Pilot operation or location installation

Phased operation or installation

Direct cutover or installation

Changing over from the old system to a new one by turning off the old system when the new system is turned on

Considered riskiest

Problems can result from data situations that were not tested or anticipated or from errors caused by users or operators

A system can also encounter difficulties because live data typically occurs in much larger volumes

Detecting minor errors is also more difficult because users cannot verify current output by comparing it to output from the old system

Companies often choose this method for implementing COTS, because they feel that COTS involve less risk of total failure

For in-house, many orgs use direct cutover only for noncritical situations

Timing issues

Direct: considered riskiest, although in my experience, it generally wasn’t

After detecting bullet point: major errors can cause a system process to terminate abnormally. With direct cutover, reverting to old system as a backup option is not possible

Timing is very important when using direct cutover. Most systems work on weekly, monthly, quarterly, and yearly cycles

When a cyclical information system is implemented in middle of any cycle, complete processing for the full cycle requires information from BOTH old AND new systems.

To minimize risk from this situation – cyclical info systems usually are converted using the direct cutover method at beginning of quarter, calendar year, or fiscal year. Often done this way at end of fiscally (crop) year at Savannah Foods

5

Direct cutover or installation

Parallel operation or installation

Running the old information system and the new one at the same time until management decides the old system can be turned off

Data is input into both systems, and output generated by new system is compared with equivalent output from old system

Obvious advantage: lower risk.

Disadvantage: primary is that it is most costly

Not practical if old & new systems are incompatible technically, or if operating environment cannot support both systems

Also inappropriate when the two systems perform different functions or if new system involves a new method of business operations

After data is input: when users, management, and IT group are satisfied that new system operates correctly, old system is terminated

Advantages:

If new system doesn’t work correctly, company can use old system as a backup until appropriate changes are made

Much easier to verify new system is working properly under parallel operations than direct cutover, b/c output from both systems is compared & verified during parallel operation

Disadvantages:

Because both systems are running in full operation at same time, company pays for both systems during that period

Users must work in both systems & company might need temporary employees to handle extra workload

Running both systems might place a burden on the operating environment & cause processing delays

7

Parallel operation or installation

Pilot (or location) operation or installation

Trying out an information system at one site and using the experience to decide if and how the new system should be deployed throughout the organization

For example, a new sales reporting system might be implemented in only one branch office, or a new payroll system might be installed in only one department

The group using the new system first is called the pilot site

During pilot operation, old system continues to operate for entire organization, including pilot site

After system proves successful at pilot site, it’s implemented in rest of org, usually via direct-cutover method

After 5th bullet point: consequently, pilot operation is combination of parallel and direct cutover methods

Restricting implementation to pilot site reduces risk of system failure

Operating both systems for only pilot site is less expensive than a parallel operation for entire company

If a parallel approach to complete implementation is used later on, changeover period can be much shorter if system proves successful at pilot site

9

Pilot (or location) operation or installation

Phased operation or installation

Changing from the old information system to the new one incrementally, starting with one or a few functional components and then gradually extending the installation to cover the whole new system

Implemented in stages

Example: instead of implementing a new manufacturing system all at once, the materials management subsystem is installed first, then the production control subsystem, then the job cost subsystem, etc.

Each subsystem can be implemented by using any of the other three changeover methods

Advantage: risk of errors or failures is limited to the implemented module only

Less expensive than full parallel operation b/c analyst has to work with only one part of system at a time

Disadvantage: can’t be used, however, if system can’t be separated into logical modules or segments.

Analysts sometimes confuse phased and pilot operation methods. Both methods combine direct cutover & parallel operation to reduce risks and costs

With phased operation, however, only a part of system is given to all users, while pilot operation provides the entire system, but to only some users

Advantage: if a new production control subsystem fails to operate properly, that failure may not affect the new purchasing subsystem or the existing shop floor control subsystem

Disadvantage: if system involves a large number of separate phases, phased operation can cost more than a pilot approach

11

Phased operation or installation

Relative risk & cost characteristics of 4 changeover methods

13

Other issues

Operational and test environment

Test environment has already been mentioned

Actual system operation is performed in the operational environment, ideally on a separate machine (mainframe, midrange, server) than the test environment

If you cannot do this on two separate machines (in order to ensure the integrity of the programs and the data), you are going to have to be able to partition your machine in such a way as to guarantee the integrity

Access to operational environment is limited to users and must strictly e controlled.

Test environment contains all copies of programs, procedures, and test data files

Access to operational environment – systems analysts, developers, programmers, should not have access to this environment for fear that they will contaminate production programs and systems

Test data files may be complete copies of production files or subsets of them.

15

Training plan

Should be considered early in systems development process

As documentation is created, consider how to use the material in future training sessions

When system is implemented (or right before cutover), consider when and where the training will take place, and which employees will receive the training

Org must consider how to support business operations while training is being conducted

Three main groups for training are users, managers, and IT staff (yes, there may be IT staff who were NOT involved in the creation of the new system / conversion from old to new)

A manager doesn’t need to understand every submenu or feature, but he or she does need a system overview to ensure that users are being trained properly & are using the system correctly

Users need to know how to perform their day-to-day job functions, but they don’t need to know how the company allocates system operational charges among user departments

IT staff people probably need the most information.

IT personnel: to support the system, they must have clear understanding of how system functions, how it supports business requirements, and skills that users need to operate the system and perform their tasks

16

Training topics for different audiences

17

Vendor training

ERP or COTS software – the vendor will probably supply it.

Gives best return on training $, because it’s focused on products the vendor developed

Scope of vendor training, however, is usually limited to a standard version of vendor’s software or hardware. Problem with this?

Problem is that if you modified the software or hardware with the vendor’s permission, they may not be able to train you on it.

18

Training tips

Train people in groups, with separate training for distinct groups

Select most effective place to conduct training – at company’s location has several advantages:

No travel $$ incurred

Can respond to local emergencies that require immediate attention

Can take place in actual environment where system will operate

Disadvantages at company location

Employees who are distracted by telephone calls and other duties will not get full benefit

Provide for learning by hearing, seeing, and doing. Not just lectures

Rely on previous trainees – train the trainer – now the trainer is an employee who knows the system

questions

You are installing vendor-supplied Enterprise Resource Planning (ERP) software. The vendor is in Atlanta. Should you have the vendor come here to Savannah, or send your employees to Atlanta?

Your company has multiple locations that are all connected to one machine at corporate HQ. You are going to install that ERP software. How should you do it?

20