Feeds

back to article Performance monitoring is Someone Else's Problem

Douglas Adams obviously knew what makes an IT shop tick. In Life, the Universe, and Everything, he identified the Somebody Else’s Problem (SEP) field, which renders some things not so much invisible as unnoticeable. For a while, the imminent collapse of the Greek economy was an SEP, until it became too big to ignore. IT …

COMMENTS

This topic is closed for new posts.
Silver badge

Try SNMP

Simply Not My Problem

0
1
Headmaster

"an SEP"

My English grammar is bad - but surely that's not correct?

1
0
Bronze badge
Headmaster

Phonetics not being my strong point .....

The 'S' of SEP would take the sound 'es', this is a vowel sound. Stet.

1
0
Bronze badge
Headmaster

Indeed

I believe the use of the indefinite article is either 'A' or 'a' where the word following doesn't start with a vowel sound. 'An' is used where the next word starts with either a vowel, or takes a vowel sound. For example, a Microsoft project manager, an MS project manager (they are both of course an oxymoron anyway).

0
0

apparently, I've learned...

a new rule today!

0
1
Silver badge

SEP? Not forgetting it's close cousin

Bistromathics

The sort of mathematical rules that come into play when you try to divide up a restaurant bill. No matter how you do it, it never tallies with the amount on the invoice.

The problem with performance management is that no matter how you do it, it always fails. You monitor all your services. Identify a bottleneck. Spend ££££'s to fix it. Sit back in the glow of a job well done. The phone rings and it's users complaining about the NEXT bottleneck, now that the original one has been relieved.

Bottlenecks are like traffic lights: as soon as you get past one set of delays, you get a little further and the next one gets you.

So the net gain of the ££££'s spent is a small, imperceptible and soon forgotten benefit - whereas the cost is a monkey on your back forever. Each time you ask for more money to fix a performance problem, the bean counters remind your boss that the last attempt didn't work, or was only effective for a few sort weeks. Even if you have the experience to say "ah ha! we need to fix not only the prima-face problem, but all the structural issues behind it" and propose a cost-case to do it, you usually find that the problem goes so deep, the costs are so high and the upheaval so intense that you don't stand a chance of getting it approved. Certainly not for the miniscule and intangible benefits you can only _estimate_ it will bring.

Even invoking the third law of project proposals: The higher the price, the greater the chance of success.

Experience has shown that the best way to deal with performance problems is to ignore them. Leave them until they either cause something vital to crash and burn OR that they start to affect the CEO's computer. (In that case, fix his/her's machine and maybe "have a quiet word" while you're in their presence.) However, to succeed in this strategy, it's vitally important that you do not have any responsibility for systems performance, capacity planning, service quality or any of the other buzzwords that could let someone legitimately ask "Why did you let this happen?". Provided you leave it long enough, and the performance melt-down is severe enough (and can be shown to be someone-else's fault) you can get practically anything you like to fix it - except, of course, a raise.

Just as with the restaurant and settling the bill, performance is something everybody has, but nobody wants to pay for.

4
0
Facepalm

funny

Eradicate SEP fields... yeah right, that'll never happen.

What happens in the real world is companies struggle and bungle their way forward and getting some internal service done depends on how much clout you can bring to bear in a "my dad can fight your dad" kinda way.

Then at the end of their tether some business leader will sell out the company systems to the latest carpet bagger from "we'll sell you the solution industries inc" . This usually causes more trouble than it solves and the cycle starts again.

Eventually, that company fails due to market disruption by other companies with more of a clue.

1
0
FAIL

Leveraging??

SEP is positively encouraged by companies that 'silo' support teams using the ultra fashionable leveraged support model. It's not that the silo'd teams want to become SEP orientated, it's instigated from the upper management down who read too much junk from 'research' companies who promote 'industry best practices'. For 'best' you should actually read 'cheapest'.

0
0
Silver badge
Joke

The question is...

where did all these IT departments get all this pink paint?

1
0
Stop

Nothing new in IT

In the IT sector one is generally hired for a reasonably specialised purpose, like administering databases or servers, or coding, or architecting systems, or some managerial or directorial position that's responsible for overseeing the integration of these functions of business. The same is true of non-IT businesses; human resources, marketing and accounts departments each have their own inputs and outputs, and expect to work with only a brief overview of the business' objectives and a few lines of communication between each. Each silo has people with specialist skills who need

The 'culture of no' is merely a demarcation of responsibility. It's up to someone higher up to splice business objectives into a series of connected objectives to enable the capacity an organisation - or in this case - a system needs in order to function optimally. If a component of a business is failing to deliver, be it internal or an external dependency, then identify the culprit and pursue remedies or alternatives, rather than blaming underresourced specialists. That's good management... and rare reality.

0
0
This topic is closed for new posts.