Problem Management

Practical Case

The Service Desk of "Cater Matters" has informed Problem Management about an incident which could not be associated with a known error and which caused a low impact interruption to service.

Problem Management decided to analyse the problem following the established protocol, which is based on the Kepner-Tregoe method:

• Identifying the problem.
• Classifying of the problem.
• Establishing the possible causes.
• Checking the most likely cause.
• Confirming the actual cause.

Identification: In the case with which we are concerned, the problem is easy to define:

• The online orders application produces unpredictable errors when recording certain orders. There is no apparent relationship between the error and other hardware/software components.

Classification: The problem may be classified according to the following parameters:

• Identification: Problems recording orders.
• Source: Online orders module.
• Frequency: the problem is not recurrent, this is the first time it has been detected.
• Impact: slight. The incident was resolved without a serious interruption to service.

Possible causes: The most likely causes include:

• Errors in programming on the client side of the application.
• Errors in the web server recording modules.
• Database configuration errors.

The analysts decide that the most likely origin of the problem is in the application's recording modules.

Checking the most likely cause: with the help of the information recorded by Incident Management:

• Problem management tries to reproduce the problem.
• They find that the error is only reproduced with a particular brand of ice-cream.
• They notice that the brand of ice-cream has an apostrophe in its name and that if this is removed the order is recorded without problems.

Verification:

• A test environment is set up reproducing the module of interest on the live environment.
• The necessary programming changes are made.
• They confirm that the order is recorded correctly.

The problem has been converted into a known error. It is now the task of Error Control to:

• Raise an RFC with the proposed solution.
• Carry out the post-implementation review if Change Management considers it appropriate to implement the RFC.