Page 1 of 1

When to reopen an existing closed problem case

Posted: Mon Oct 17, 2011 3:37 am
by kamran
I would like to know, when can we reopen an existing problem case, closed with a permanent solution.

e.g. there was an outage for Service A . After investigation, a root cause was identified and a permanent solution was implemented.

After sometime, the same Service A was affected, but at this stage no initial investigation was carried out to identify the root cause.

Now at this point, should we open a new problem case to investigate to cause of the 2nd outage or reopen the 1st problem case.

Posted: Mon Oct 17, 2011 5:02 am
by Diarmid
Two answers:

Practical:
How will re-opening a closed problem help in the resolution?

Theoretical;
If you have not done an investigation into the underlying cause of the new outage, how do you know that it is the same problem? - symptoms can be misleading.

Third answer:
If they are the same problem (i.e. your permanent solution implemented was in fact inadequate), how much of the old problem process do you unwind? Did it go wrong in the analysis or solution part (or somewhere else)? Why would you re-open a flawed process?

Posted: Mon Oct 17, 2011 6:11 am
by kamran
@ Diarmid,

Well it is well explained and understood, thanks for that.

But what would I do in case where after the investigation it turns out to be the same root cause as it was for the previous closed case?

or

if I know for a fact that the underlying cause is the same as it was for the previous closed case.

Is there a good practice or criterion to check against to reopen a closed case, outlined by ITIL or we must open a new case and relate the old case with the new one to give it more weight and ask the service owner/proponent to do a proper assessment and propose solutions to prevent from happening again.


or it all depends on human judgement.

Posted: Mon Oct 17, 2011 9:58 am
by Diarmid
I cannot envisage any circumstance in which it would be best to re-open a closed problem. Certainly it makes sense to reference the earlier investigation, but what reason can you propose for re-opening it?

I would attach a note to the earlier work stating that it failed and why it failed and have your quality manager or problem manager attack the causes of the failure. I might even raise a problem about the failure. These things would be harder to do if you re-opened the case, because there would be a new final outcome.

It's a different matter if the new incident occurs while the old problem is not fully closed, for example, pending routine review and final sign-off.

Posted: Wed Oct 19, 2011 2:25 pm
by Sunny60in
Hi Kamran,

Your question "... should we open a new problem case to investigate to cause of the 2nd outage or reopen the 1st problem case."

To me it means... It is not yet confirmed that root cause is same at case 1. So this is good reason to open a new problem. You might end up with some differnt root cause. I am sure investigation would start 'after' problem case is opened. If root cause turns out to be the same, link these problems.

So... start with new problem link if outage & root cause are same.

Posted: Thu Oct 20, 2011 2:23 pm
by Diarmid
Why would you link the problems?

Posted: Thu Oct 20, 2011 4:29 pm
by Sunny60in
I suggested to link problem because problem & root cause were same for the same service. I have never come across similar situation though. May be problems shouldn't be linked. Please explain if it not a good practice.

Posted: Fri Oct 21, 2011 9:50 am
by Diarmid
It should be such a rare occurrence that there would never be a use for the link. I can't imagine needing a report of problems that needed to be dealt with twice. You might want to put a note in each one just in case someone needs to look at the history sometime.

On the other hand your quality manager would want to see that someone had investigated why a problem had been closed with a false/inadequate resolution in case it indicated a failing somewhere in the management system (poor diagnostic skills, poor testing, poor controls?). This could be done in the problem review when it is finally closed or in the regular problem management review or as a special review of the occurrence.

And if it did ever happen again (to any part of the service) there definitely ought to be a review. Getting as far as investigating a problem and applying a solution only to find it has not gone away is likely to be costly and to damage confidence in the whole of service management.

It's not something you expect to happen, and so putting a link in seems a bit of a cop-out, as if to say "this might be interesting to someone someday", instead of "this needs looking at now".

Of course there are other circumstances when it is relevant to link problems, such as linking all problems related to a particular service or a particular supplier's hardware. But your system or your routine reports may well make those links anyway.

Posted: Wed Oct 26, 2011 1:27 am
by smehdi
I think that the a new PR should be open for a new incident as at that moment we dont the root cause of the issue.

Once the root cause is confimed and you found that its same as the previous one , plz go ahead and link it to your previous PR for the trending purpose.

I normally raise a new PR for every incident requires investigation and link it to th existing PR if the root cause is same (Parent /Child PR concept)

Additionally , if you open the existing PR , the investigating team might get influenced with the previous findings in the PR and could lead to same result. However a fresh PR might give them a new direction of investigation.

Posted: Wed Oct 26, 2011 5:35 am
by Diarmid
smehdi wrote:Once the root cause is confimed and you found that its same as the previous one , plz go ahead and link it to your previous PR for the trending purpose.
If you are trending the recurrence of the same problem you are not doing very well. It should not happen! Don't forget kamran is talking about a problem believed to be already resolved.