Troubleshooting Methodology and Practices - SQL Server 2012 Internals and Troubleshooting - page 244

9.

Test and implement — Even if the problem does not exist in the test or pre-production
environment, implement the i x there i rst. This involves making the identii ed change and
coni rming no undesired impact, then deploying to the production environment. If possible,
ensure a rollback position and be prepared to invoke this plan if necessary.

10.

Review — Post-mortem analysis will help prevent further recurrence of this issue or new
issues in the future and can be used to identify other vulnerable systems within the organi-
zation which should be i xed, and will improve the troubleshooting approach to ensure it is
as optimized and efi cient as possible.

The ten steps outlined above and described in more detail in the following sections describe a trou-
bleshooting approach you can adapt and simplify as desired. Not all problems require full formal
engagement, but adopting an approximation of these disciplines can help you prioritize other activi-
ties, such as monetizing the impact of problems and dei ning a clear problem statement.

Behavior and Attitude
In addition to employing a good troubleshooting approach, adopting a positive attitude with mod-
erate determination and persistence to identify the root cause and resolve issues dei nitely helps. A
positive attitude leads to better quality results, faster resolution, and it will reduce the stress level
for you and co-workers during the troubleshooting process. Using a consistent approach to resolving
problems by decomposing them scientii cally is a proven and effective method, and many of these
aspects are within your control.

The following behaviors and attitudes are characteristic of the most effective database professionals
when troubleshooting complex problems:

➤

Remain calm — Stay objective, no matter how urgent the problem. Project coni dence and
calmness to your peers, end users, and management, even if they show signs of stress or
panic. This reassures them that you are in control and able to resolve the problem. These
people are more likely to give you the time and space necessary to investigate and resolve the
issue if they trust your capability.

➤

Remember that problems are never random — Problems with computers happen for a
reason. When you don't understand the reason, the cause may seem random, but there is
always an explanation. Intermittent or infrequent problems in particular appear random;
seek to identify patterns or correlating events that could lead to the circumstances that cause
the problem.

➤

Avoid prejudice — Never assume that you know how to solve a problem until you have
a problem description and have done some basic testing. It is not necessary to provide an
instant answer; the correct answer with a short delay trumps a quick, inaccurate answer.
This habit also builds your credibility with management as a reliable and capable engineer.

➤

Avoid looking for i xes — Ensure that i nding the cause is your i rst priority! The people
around you will be pressing hard for a i x or an estimated time to i x. The i x is the goal,
but you must i rst lay the foundation by understanding the cause.

➤

Think ahead — Proactively consider potential blockers. If you may need to restore the data-
base, start the tape retrieval process in parallel with troubleshooting. This reduces overall
downtime and impact if you do need to revert to the backup.