In many organisations the costs of data warehousing are expressed on the balance sheet as the hardware, software and administration costs that support the operation. In practice, many hidden costs are associated with the data warehouse with the structure of an organisation. So says Mark Thomas, IBM’s Infosphere and data …
One of the big problems in this whole area can be ownership of the entire process. One team is responsible for infrastructure, another for development - with an offshore team who write masses of very simple scripts / ETL processes which do nothing more than mirror existing data holdings onto tables on the Warehouse. Then you have a team who advise on strategy, but often they are ignored. And service managers, but they are largely absent.
So you end up with masses of duplication, which, although cheap to initially develop and deploy, bring the infrastructure to it's knees. So the infrastructure team has to grow to cope with all the incidents, upgrades and so on.
Crucially, if the data itself is not treated as a business asset, owned by a central team, there is no interest from each business strand of going through the pain of sharing or integrating into a common data model.
Spot on Sir, basicly this is a story about accountancy failure to adapt to the real IT world and impose a standard of accountancy that does that place true value upon the data that is managed and respective returns.
IT department costs it's services at cost to the internal departments and as such never get shown to provide a return to the company. Now if said department saves half it's staffing costs and makes alot of money thru there saving and more effecient ability to analyse the market then there department shows a profit without any of that return factored down. In effect all IT is costed and accounted as a fixed cost with any overrun/failure in any shape of form accounted against the IT department and and return and profit accounted against the non IT department. These are both departments working for the same company and they wonder were the team spirit is!
"Too many cocks spoil the data warehouse brothel..."
Gotta lay off the 4chan...
Bizarre sales pitch.
Firstly he seams to be adding the users time into the IT costs.
Secondly he cites a number of problems most of which are symptoms of poor management and lax governance.
Then he proposes the "solution" which is to buy another hardware platform from IBM. How managing yet another piece of hardware and yet another software stack solves the problem of bad management I fail to see.
Although targeting dumb managers is probably good for his bonus.
Not if you know IBM
IBM loves to sell services/or a package with managed services as apposed to just a box. Now they will happily price up a perfect datawarehosue solution for you incuding running the service. Some accountant who don't fully understand IT will look at the service and say we can get X staff for that cost and do it inhouse. They call the IT department say can you do this and they will cost it up for extra staff, initial training etc as they welcome any bone to grow there departments, all department heads do nomatter what department they run it seems. They then find that the difference of IBM managing it and running to the spec IBM outlined have differences. The differences being that IBM are more able to sort out any issues behind the scene without you knowing about it and have the resources there and ready to call up who built the software etc as apposed to your IT department that whilst will have tallented staff who know more skill sets than there paid for and yet still have to call IBM to resolve deep down issues such as quirk with the DB2 configuration or other aspect that is so rare your IT department hasn't seen it before and nor has google and remarkably enough the people who wrote it know all about. Not saying IBM the best, its just a situation alot of companies get themselfs into be it IBM or any other software. Once you start doing your own support, like opening the back of your iPhone you start to blur the lines of responsibility and as such responsibility and the accountability of who pays for the impact to service and also the response to that incident in time can esculate the visability of the incident. When you do it inhouse, one light failing on a blinking led panel becomes a known issue and concern, if it's supplier managed as a service then you know nothing about that and never will.
If any problem in IT overrun's it is always down to being poorly specified and or costed wrong. What alot forget is once you get a cost for X solution and support and then break that support out and do it inhouse that the cost is relative in that your supplier can easily and cheaply do alot of things that will cost you inhouse more and that if you pay the same level of money in some area's you might realise that the supplier's cost is not the golden value you were looking for. That is poorly specifying the requirements from the start. Get a costed solution for your problem stating you will want to run this inhouse and supported inhouse initialy and then get the services cost price, that way around you will get a more true picture - but you need to specify what you realy want at the start or you might as well piss in the air and work out the wind direction that way.
the number of sysads to a project is inversly proportional to its efficiency
Name me one DBA team that has
I've google'd, wiki'd, interwebbed and I have yet to find a DBA team that has lost many many many billions of real money one year and then goten lifetime salary bonus's the next year.
Not even bank DBA's have achieved this!