Data Fix Process

General discussion on all aspects of the IT Infrastructure Library (ITIL)
Post Reply
User avatar
DavidStocks
Newbie
Newbie
Posts: 1
Joined: Mon Sep 29, 2014 8:00 pm

Tue Sep 30, 2014 11:38 am

Hi All,

I work on a fair new helpdesk and we are currently reviewing our processes around data-fixes.

To give you a bit of the back story, an member of the team ran a data-fix and accidently delete an entire table in SQL.

I know that in theory, you are always supposed to take a back up of any changes made. However (in this example)One value on one row needed to be updated for a table that holds close to a million entries.

There has been the suggestion that the entire database should be backed up or that the entire affected table should be backed up. Like I say in theory both sound perfect but in practice its not always feasible to take a full back up before such a minor database. Retrieving a full back up from our clients can often take days if not longer.

I'm am wondering how everyone else copes with these types of issues?


User avatar
UKVIKING
ITIL Expert
ITIL Expert
Posts: 3639
Joined: Fri Sep 15, 2006 8:00 pm
Location: London, UK

Tue Sep 30, 2014 2:25 pm

In my org, we treat back end data changes like application code changes

1 - The script is run in non production to validate the execution
2 - the deployment to non production is not done by the development team
3 - then the solution is then done to production

now in addition, we determine the criteria for changes that need full CM approval
John Hardesty
ITSM Manager's Certificate (Red Badge)

Change Management is POWER & CONTROL. /....evil laughter
User avatar
thannguyentan
Itiler
Itiler
Posts: 5
Joined: Mon Oct 06, 2014 8:00 pm

Wed Oct 08, 2014 12:13 am

I think you need review your backup & restore process again
- When we need backup full
- When we need backup incremental
- When we need backup daily
- When we need backup weekly

And this process must be writen by your Technical leaders and Database admin, they are expert in this field.
- Verify process before we can trust this by running test case.
- if there is any problem in process must be logged on Problem management.
+ Scenario
+ The way of backup
+ What is the issue
+ where is the root cause
+ How to resolve
+ Who is expert engineer....
+ .....

We should analysis and review backup process by testing our process.
Beside that we need to build audit plan to assess all process to ensure process up to date. And can apply for our current environment.

Just by my thinking
Post Reply