Table of Contents
A problem is a very useful tool in continuously improving the quality of your service. Whenever a series of incidents of the same type appears a problem should be logged to investigate. The problem handling process is aimed at finding the root cause of related incidents and finding a permanent solution. Even if a permanent solution cannot be found right away, the problem can be logged as a known issue and a work-around can be provided with the known issue to make sure incidents can be handled more promptly.
The process of problem handling consists of the following steps: Reporting a problem and linking all related incidents to the problem, investigating the root cause, find a solution and/or work-around, report back in all related incidents, and close the problem.
When a support engineer detects that similar incidents, or at least incidents that have the same probable cause, he or she can choose to report a problem. When the problem is logged, the following information should be provided (if known):
Symptoms of the incident as reported
Affected services
Steps to reproduce
Work-around(s)
Detailed description of what has been done to (try to) fix the incident(s)
Impact to the user (how does it affect their use of the service)
When this information has been gathered, all related incidents should be found and linked to the problem. If the incident is already closed it may still prove useful as the problem is being investigated, therefore, recently closed incidents should be added as well.
Now the problem can be assigned to someone that will be investigating the problem. This should be someone with detailed knowledge of the affected services and someone that has sufficient time to investigate the problem in detail.