Wednesday, December 21, 2011

IT Production – Cinderella or Ugly Sister?

IT Production – Cinderella or Ugly Sister?

With the pantomime season fast upon us, many IT managers will be dreading the pager call in the middle of Act 2 “IT’s behind you!”, or rather “IT’s not working”. While the rest of us bask in the warm friendship of friends and family, other poor lost souls may well be struggling with a callout on a cold hardware failure, or a stubborn Oracle or SAP system refusing to produce the correct results.

For others, it will be a case of “I can’t come to the ball (for which read: consume massive quantities of alcohol etc. etc.), I’m on call this evening.”

Like it or not, our information-industrial society cannot survive without 24 * 7 IT. And that means 24 * 7 IT Production Support.

Surprisingly, given that for many companies as much as 75% of their IT Budget is spent on Business As Usual (“BAU”) or Production issues, it seems there are too few answers to the question of how to best manage this vital part of the IT landscape.

One approach to addressing these challanges is “MOPS”.

As is common these days, MOPS is a 4-letter acronym. (At which point, it may be worth remarking that there has been a plethora of FLAs – or four-letter acronyms – over the last few years. This is surely another indication of the growth of IT exceeding its “name space”).

The “M” in MOPS stands for “Metrics” and highlights the importance of capturing meaningful data on the Assets and Activity of IT Production. This can include all sorts of data from an Asset Register or small Configuration Management Database (CMDB) to track the growing responsibilities of Production. Also included are Key Performance Indicators to show Service Levels, Incidents and Callouts. Not to be forgotten is the importance of capturing the activity of the Support Teams by means of Timesheets. All of these metrics, properly collected, managed and structured, can go a long way towards helping the IT Production manger with his biggest challenge – explaining and justifying the IT Production costs.

Operational Tools comprises the second letter. Simply put, this part of the acronym highlights the importance of Software to help run Software and Hardware. The range of useful software extends from Backup utilities, Software Distribution, Monitoring, Management, Alerting and Capacity Planning. However, Software Tools in and of themselves would be unable to deliver the levels of efficiency and automation in IT Production unless they are properly integrated. In fact, the cost of IT Management Software is relatively cheap. The real cost of deploying such tools is the implementation and integration within the IT Production organisation. Here, the MOPS approach discusses how to establish a Monitoring and Management Tools Strategy, and how to build a “Referential” approach to Software Tools integration.

As we all know, effective management of IT is impossible without clear and appropriate Processes and Procedures –the “P” of the MOPS acronym. This is where the ITIL framework dovetails into the MOPS approach, since MOPS recommends using ITIL as the basis for defining processes. However, MOPS also identifies other processes such as how to manage Infrastructure Research & Development, how to build engagement processes with the Development projects, and how to build a Standards Governance process. MOPS also recommends using process improvement techniques such as Six-Sigma and LEAN to increase the efficiency of existing process workflows.

Often forgotten in IT Production, the final letter of the acronym (“S”) refers to the importance of IT Production Standards. Here, the MOPS approach discusses the importance of defining Infrastructure Technical Standards and how to define an IT Production Strategy and Architecture role to act as “gatekeeper” to the production estate. How many times in IT Production have we seen applications deployed that are functionally rich from a business perspective, but lacking in supportability, without reference to the IT Production imperatives such as resilience, scalability, backup and recovery capability and the ability to be monitored in Production. MOPS seeks to address this by looking at the definition of “Production supportable”, and setting a Standards process in place to be applied to proposed new application developments.

So MOPS may be one approach that could enable us to get a better level of control over the frenetic world that is IT Production.

So will things change in future? I hope so.

For a long time, the IT Production team was the Cinderella of the story. She was kept away in the dark recesses of the organisation, with few people realising or acknowledging that she was a vital part of the story.

Armed with “MOPS”, she may be able to sweep out some of the cobwebs that have built up over the years, and even be able to take time off for a party or two. Who knows?

I wish you all well in the festive season, whether or not you are on callout.

Thursday, November 24, 2011

Where has IT Quality disappeared?

What happened to the concept of Quality?

In the old days of IT (or Computer Science, as I believe it was once called), the main practitioners were from an engineering background or civil service background. The idea of procedures and processes were pre-eminent.
Such was the importance of computing power that detailed design, walkthrough and analysis was used before people were let loose to cut code, or to implement as system.
This, in turn, led to the importance of processes and controls, and to the so-called Mainframe mentality. It was alleged that it took years for anything to be done on a mainframe, which was one of the justifications for the later Client-Server revolution (remember that?) and the rise of midrange systems (remember them?) which were said to be easier to deploy and implement.
Despite a period when the term Software Engineering was used, the cultural norms have moved towards the world of Agile Programming, Extreme Programming, Prototyping etc.
In the world of Infrastructure, there has been a rush to catch up as the Infrastructure desperately tries to keep pace with the rush of new Applications being thrown over the wall onto it.
In the meantime, we now have chaotic organisations, where management processes (control of change, scheduling of activities, etc.) are equally Agile (or disorganised, depending upon your point of view).
So where are we now?
Some organisations I have seen have Chaotic Programming, Chaotic Infrastructure Management and Chaotic Team and Processes.
Is this the best way to run Information Technology?
Sometimes I think not.
Yes, we must be agile and responsive to the business.
But if we cave in to the pressures for unstructured and ungoverned change we will be creating issues for those that follow us in the IT world.
For many, who live as contractors, this is not an issue; the next contract is around the corner, and someone else can pick up the pieces. Even CIOs are not immune to this short-termism.
Maybe it's time to take a longer term, more responsible view of the way we do things?