ITIL Alligators

CanStockPhoto kovalvsHere's a simple analogy for incident, problem and change that I use quite a bit.

"When you are up to your arse in alligators it's hard to drain the swamp".
[Updated 16/9/09:]

Alligator bite=incident
Band aid=incident management
Tourniquet=major incident management
Alligator shooting=problem management
Fence around the swamp=change management
Gate in the fence=release management
Hole in the fence too small for alligators=standard change
Drain the swamp, remove the environment in which alligators thrive=continual improvement
Alligator's nest of eggs=project
Hatching time=production handover
Alligator smuggler=cowboy tech
Map of swamp=configuration management
Alligator tagging=known error database
Problem Manager wanted - apply within: CanStockPhoto CCarvellProblem Manager wanted - apply within CanStockPhoto CCarvell


Drain the swamp

I've always thought that the traditional saying:

"When you're up to your ar** in alligators it helps to remember you're there to drain the swamp"

to be a pretty dumb one. It just pushes back to the question of why you want to drain the swamp?

Well of course you want to drain the swamp so you can have some solid land as beach-front in Florida so you can build a golf course for the condominium retirement home for rich retired stock brokers so you can make piles of money and go live on a yacht in the Bahamas.

So when we go on about the details of fencing, nesting, gates and all that, we're just reverting back to techie-geeks and forgetting that the real business of business is to do business and make a profit. Never ever forget the economic motive.

ecological activists

When you want to try to build something in IT, often the IT techs are the ecological activists, up in arms over the destruction of the habitat of their beloved alligators.

We hire techs because they love to solve problems. it is only one step in attitude to love problems. How many techs do you know who relish problems? Polish them and show them to everyone before they reluctantly put them down bacuse that's their job. After they do, they preserve them in a jar, telling folk about them for ages afterwards.

I love it, but..

LOL. No, hold on, I never say that. LOLgators?

Alligators aren't the problem. They've been around for thousands of years. The problem is when alligators and people have to co-exist. Both of us have legitimate aims.
And alligator shooting doesn't solve the problem, in fact it could cause a lot more unexpected incidents - I would say that was major incident management.

Problem management has more dimensions to it.

Standard change = relocating rogue 'gator to a new location?

Change management has to be multi dimensional - change one element and you get a sub optimal solution?

Draining swamp etc. misses the point. Why do 'gators exist? Why do we want to eradicate them? Wouldn't life be easy if we eradicated business driven changes?

Question: Who are the 'gators?

Does the charging'gator have to bite you before you take action?

I can't help but think that 'real, perceived or threatened deviation from planned and agreed service behaviour' would align better with the customer/users expectation of something that merits creation of a Service Incident (SI). Thus don't wait to be bitten before starting to manage the situation with the appropriate urgency. If for example 1 of 2 clustered nodes has failed then manage the SI according to whether you still have 100% capacity to meet the expected service level or not and what service impact you would have if the second node fails before the first is restored. Managing the associated but separate Problem(s) of why the node failed might be done to a different tempo.

Dave Schofield (Sheep dipped and in no great rush to be sheared)

Dave, being sheared is okay,


being sheared is okay, its getting fleeced you want to avoid :-)

if a 'gator is getting too close

You're falling into the same trap Dave. "The only way to manage the situation with the appropriate urgency is via an Incident". No it isn't. there is an entirely different process if a 'gator is getting too close than there is if it has already bitten you. in the same way, the process to resolve a fault is different to the process to resolve an incident for all the reasons outlined here, but it is addressed with urgency.

Incident/not-incident does not imply urgent/not-urgent


"there is an entirely different process if a 'gator is getting too close than there is if it has already bitten you."

Perhaps proactive Problem Management?


proactive problem management

Good point. reactive problem management is shooting a gator that has shown itself to be a man-attacker. That common response of going out to shoot a maneater once someone is dead. To be controversial, I also categorise incident matching and other trend analysis as reactive problem management not proactive. You're still using bites as evidence.

proactive problem management is several things:
- picking the buggers off before they hurt anyone: risk management, having people auditing systems for potential problems, then fixing them as part of Availability Management and/or CSI
- crating them up and moving them back outside the fence: e.g. demoting systems from production-controlled to un-controlled, refusing to accept substandard projects
- advising on swamp draining and otherwise making the swamp unwelcoming to gators: contribute needed reforms to CSI planning


I tend to your point 1. I would sit on the porch and shoot every gator that showed himself.


Shooting gators

The difficulty with that plan is that IT's alligators breed like rats and are very hard to kill, we don't have the equivalent of a gun. Each one takes time and money, there are lots and lots of them, and some of them live in the swamp all their lives without biting anyone. Many of them never grow big enough to be a threat. You need to pick and choose your targets


I could imagine a whole book/training course/philosophy base around this idea.

Syndicate content