5.1 - How To Troubleshoot

Question 1

Change management

Answer 1

• Change control
  – A formal process for managing change
  – Avoid downtime, confusion, and mistakes
  – Corporate policy and procedures
• Nothing changes without the process
  – Plan for a change
  – Estimate the risk associated with the change
  – Have a recovery plan if the change doesn’t work
  – Test before making the change
  – Document all of this and get approval
  – Make the change

Question 2

1. Identify the problem

Answer 2

• Information gathering
  – Get as many details as possible
  – Duplicate the issue, if possible
• Identify symptoms
  – May be more than a single symptom
• Question users
  – Your best source of details
• Determine if anything has changed
  – Who’s in the wiring closet?
• Approach multiple problems individually
  – Break problems into smaller pieces
• Backup everything
  – You’re going to make some changes
  – You should always have a rollback plan
• What else has changed?
  – The user may not be aware
  – Environmental changes
  – Infrastructure changes
• There may be some clues
  – Check OS log files
  – Applications may have log information

Question 3

2. Establish a theory

Answer 3

• Start with the obvious - Occam’s razor applies
• Consider everything - Even the not-so-obvious
• Make a list of all possible causes
  – Start with the easy theories
  – And the least difficult to test
• Research the symptoms
  – Internal knowledge base
  – Google searches

Question 4

3. Test the theory

Answer 4

• Confirm the theory
  – Determine next steps to resolve problem
• Theory didn’t work?
  – Re-establish new theory or escalate
  – Call an expert
• The theory worked!
  – Make a plan…

Question 5

4. Create a plan of action

Answer 5

• Build the plan
  – Correct the issue with a minimum of impact
• Refer to vendor instructions
  – Knowledge base, support articles
  – Documentation and manuals
• Identify potential effects
  – Every plan can go bad
  – Have a plan B
  – And a plan C

Question 6

5. Implement the solution

Answer 6

• Fix the issue
  – Implement during the change control window
• Escalate as necessary
  – You may need help from a 3rd party

Question 7

6. Verify full system functionality

Answer 7

• It’s not fixed until it’s really fixed
  – The test should be part of your plan
  – Have your customer confirm the fix
• Implement preventative measures
  – Let’s avoid this issue in the future

Question 8

7. Document findings

Answer 8

• It’s not over until you build the knowledge base
  – Don’t lose valuable knowledge!
• What action did you take?
  – What outcome did it have?
• Consider a formal database
  – Help desk case notes
  – Searchable database

Question 9

Troubleshooting Flowchart

Answer 9