|
INFOMET 911- JUNE 2005
Almost everyone in the IT Industry knew the late Coert Vorster. He was most probably the most networked IT Executive in South Africa. During a Springbok Rugby match at Newlands in 1999, Coert was sitting behind the CIO of Old Mutual at the time, and overheard him saying to a colleague: “It is so sad; there was this one company that could have helped me, now they are just gone! They had great people, and they could really solve problems and create solutions.” “Who are you talking about?” asked his mate. “I am talking about Infomet” he declared.
For many years this tale haunted me, says Pieter Viljoen the company’s Managing Director. Infomet was truly one of the great companies pioneering formal Information Methodologies for Business and Information System problems. It got sold to the ISG to later become the IT Solutions Division of IBM, once they returned to the Country. And yes, Infomet was gone.
in recent years, more than ever before, the need for Infomet has become so emphasized, that a return of the company, its philosophies and methodologies became inevitable.
The Infomet Development Centre is a well organized and technologically equipped Software Factory environment headed by Craig Nel, and has already proved itself with its awesome development capability and technical experts. “We believe that our model of collaboration is a fundamental requirement of how IT development will be handled in especially South Africa in future”, says Nel.
Pieter Viljoen is quite vociferous about the issues and problems that have to be addressed, and the questions he is asking are very uncomfortable:
- Why are we dumbing down the corporate organizations?
- Whatever happened to State-of-the-Science approaches?
- Has management gone crazy by putting administrators in charge of complex engineering projects just because they happen to have a project management certificate?
- What the heck are the Auditors doing in the IT space, and how did they manage to hijack the ‘engineering’ work?
- Are the ‘obfuscated’ hyperlink kids simply a lost generation whose lack of formal and insufficient education, training and experience creating disastrous systems for business?
- Is anyone out there modelling and documenting anything at all anymore? And if they are, can someone please explain where they got their linguistic skills, because apart from really not having a proper command of the English language, an ability to clearly formulate and communicate concepts, they also seem to string reams of meaningless jargon and spuriously fabricated terms and phrases together which rather than impresses is simply ‘warra-warra’?
- Why are we not building the intellectual capabilities of the Business and Software Engineers? We are simply not training, mentoring and managing these resources anymore, everyone is sort of responsible to figure out their own route in the Business and IT Matrix of today.
So, Infomet is back, with the full Monty, says Lans Malherbe Business Development Director. The company is headed by a set of veterans together with of the most talented IT professionals to provide the necessary consulting, training and development services. The Infomet Methodology has been extended, and this time it really seems as if the company has its work cut out for it.
SQL SERVER 2005 IMPACTS ON DEVELOPMENT PROJECTS - MAY 2005
Infomet is running a series of readiness workshops on SQL Server 2005 in order to help managers, architects, developers and DBAs to understand the real changes that SQL Server 2005 will bring. For contact information, go to www.infomet.com
What does Infomet say about SQL Server 2005?
SQL Server 2005 is due to be released at the beginning of November 2005 and is a much anticipated release for Microsoft development shops. In addition to SQL Server 2005, Microsoft is launching Visual Studio 2005 and Biztalk Server 2006 at the same time meaning that a significant portion of Microsoft’s software development environment and architecture will hit the streets at the same time.
Just as significant is the current availability of the first public SQL Server 2005 Community Technology Previews (CTPs), allowing those who are interested in SQL Server 2005 to download the release, without all of the hassle of being on a beta programme. With a fixed release date and something you can currently review, SQL Server projects that are planned or underway must be assessed if November 2005 is within your project window.
After a five year wait one would expect something significant out of SQL Server 2005 and Microsoft has certainly done that. New features and enhancements are extensive with enough changes to the database engine and core functionality to warrant a closer look. Fancy functionality, tools, utilities and wizards will have an influence on development projects and over time will be taken up by developers, architects and DBAs until they become the de facto way of delivering solutions based on SQL Server 2005.
What is important to consider right now is not the flashy new features, but rather what features and enhancements will affect the way we are architecting and implementing our current applications based on SQL Server. There are significant changes that will mean that our application architectures will change, our development standards and approaches will change, our resources will need to be trained, some code will be totally reworked and ultimately our project plans and delivery dates will be affected. With wide ranging impacts on what we have to deliver it would be foolish not to consider the impact that the pending release will have on our projects.
The detail on new features and enhancements is readily available and will no doubt be extensively covered by Microsoft during and after their product launch but certain changes (although they may seem small) will have an influence on planned and currently active SQL Server projects; some, listed below, I believe need to be considered:
- .Net Integration: Much is going to be said about the .Net Framework CLR (COMMON Language Runtime) being hosted in the database engine whereby functions, stored procedures etc. can be written in .Net. While this does have some benefits I expect that traditional Transact-SQL will still be favoured for most database functionality due to performance considerations; however, the use of .Net components within SSIS (SQL Server Integration Services) could add a lot of flexibility when developing ETL applications.
- Transact SQL Error Handling: The addition of try-catch constructs to T-SQL is a long overdue enhancement that provides T-SQL with some maturity as a programming language. Every stored procedure will need to be re-written to take advantage of this important enhancement.
- SQL Server Integration Services: SSIS is a complete rework of DTS (Data Transformation Services) and offers a sophisticated environment for developing applications that need to Extract, Transform and Load data (ETL). Development projects where ETL functionality was previously hand-coded in external applications or T-SQL can now be developed using a completely different environment. Since many projects do ETL data, architects and project managers will need to understand the suitability of SSIS for their projects and how it will affect their implementations.
- Service Broker: The new Service Broker is basically a message queue based mechanism embedded within SQL Server 2005. While this may seem to be trivial functionality, it could have an impact in terms of the overall application architecture; as the existing Microsoft approach to message queues is to develop using .Net components and MS Message Queue, where the components themselves interface with the database. The Service Broker places queues much closer to the data itself, requiring a rethink of the architecture on some projects.
- Database Mirroring: Many manageability and availability enhancements exist in SQL Server 2005 and functionality such as database mirroring will affect the hardware choices and the deployment of databases.
The Microsoft approach and architecture is going to change with the release of SQL Server 2005 and Visual Studio 2005; and architects, decision-makers and planners need to use the preview and available documentation to understand the impact that this will have on their projects as soon as possible.
BENCHMARKING FOR IMMEDIATE BUSINESS BENEFIT - AUGUST 2005
The best approach to get business to perform as expected is to cut through all the hype and get to the weaknesses of either business or application processes. This in other words means benchmarking and incidentally due to limited resources and time, benchmarking is also the one thing organisations avoid most.
Infomet realised the headache many businesses are stuck with and developed the Benchmark in a box (BIB) solution to not only benchmark the business processes but to also to analyse the business processes.
Business might be good on the forefront but if the production systems working in the background can not process the information efficiently and fast enough, business will not go well on the forefront for very much longer. Addressing these shortcomings in a production environment is not as easy as getting the vendor in to rectify the problem. To analyse the shortcomings of a production system there will always be at least two stakeholders involved, the vendor and the client, each with their own agenda. The client wants a faster production system but feels it is unfair to be expected to pay good money again to the vendor for the same system. The vendor argues that the specifications changed and to deliver a faster system with the current specifications is outside the scope of the maintenance agreement between the client and the vendor. With each stakeholder having their own agenda a classic he said she said deadlock scenario is quickly created where neither one of the stakeholders want to accept responsibility of the current situation. The only effective solution to a problem like this is to involve a neutral 3rd party that will be able to effectively analyse the problem and produce a honest, 100% unbiased manager’s report at the end of the day that will in detail show the flaws in the production system whether it is flaws in the client’s processes and hardware or flaws in the vendor’s software. Just producing a detailed report is not enough however; the vendor must be kept up to date throughout the analysis process and should be actively engaged concerning their software architecture in order to produce an accurate analysis of their software.
Infomet specialises in the analysis of production systems that are underperforming and maintains a professional distance from both the client and the vendor to avoid alienation from either the client or the vendor. A typical real world example was when a major retail group approached Infomet as a neutral 3rd party to analyse the current SCM production system. The current system was performing adequately for almost a year but because the retail group had had such a successful business year the volume of data the production system had to process was growing so fast the system was not able to cope with the load anymore. Immediately the retail group blamed the SCM software vendor expecting the vendor to rectify the problem without any extra money being paid to the vendor. The client felt the system was barely in production for a year so why should they pay again to fix a ‘faulty’ system. The vendor wasn’t willing to take ownership of the problem arguing that the system was not at fault, claiming the retail group’s processes was not optimal.
To analyse the problem we used the Infomet methodology to analyse the client’s current production process and the client’s current hardware setup. We used the same methodology to analyse the vendor’s software. The Infomet benchmark framework was used to analyse the performance of the current production system on different hardware and with different software configurations. With the explosion of the World Wide Web information is limitless but in many cases not applicable to all situations and in some cases just plain marketing hype. Infomet eliminates the speculations and rumours by actually benchmarking the real production process by cloning the production environment on the Infomet Benchmark in a box (BIB) server. This enables us to produce benchmark results that is not only relevant to the unique production environment but is also in a format the client and vendor is familiar with. The BIB also allows us to do as much production runs as possible without being limited by the client’s own production systems that need to share resources. The BIB server also eliminates any marketing hype in the business world surrounding certain hardware and software architectures by producing hard facts relevant to the client’s unique production environment where profiled data commonly used in standard benchmarking does not.
At the end of the benchmarking exercise Infomet produced a manager’s report stating all the flaws in the current system, both on the retail group’s side and on the vendor’s side. Infomet also created a roadmap, through the collaboration between the vendor and the retail group, which detailed the approach each party will take to solve the inadequacies. Finally after months of deadlocked negotiations between the retail group and the vendor there was closure and both the retail group as well as the vendor could start working together to create a more efficient business system.
|
