How to Create an IT Disaster
Thursday, March 31, 2016 at 1:45PM
Bill Barclay

Ben Day put up a "come hither" Business Case for what is virtually a new IT system at the 14 March Special Meeting. I have only managed to to get to the Agenda in order to analyse just what he is proposing, and it is a typical piece of Day obfuscation that could well cost the Council a heap before it is implemented, and more important, it completely removes the need for any kind of competitive tendering for what is a major change - whoever has the 'inside running' has it for the duration, and on a 'cost plus' basis - money for jam!

So much of what Day has proposed in the Business Case is typical of the jargon-riddled cases that he has put up in the past - designed to confuse non-technical people with IT gobbledygook and self-serving clap-trap that leads straight back to the selected provider who is by now extremely close to the action and ready to move rapidly into the implementation phase.

Forgive me - I am a non-believer, and have been associated in the past, both inside and outside the public service  with countless IT proposals of this nature. Mistakes tend to repeat themselves over and over as Mr Stephen Joyce is currently discovering to his cost. That is not to say that this proposal Day is making does not have merit. It may well have, but for our Council to accept the 'pig in a poke' without outside demonstrably independent advice is utterly irresponsible. They are simply not equipped to make this decision that initially involves some $652K, but potentially far more as time goes on.

Interestingly, it is not until page 5 of the Business Case that the name of the ultimate beneficiary supplier is named - Network Edge NZ Ltd. This is obviously a highly qualified and competent supplier, but no one around the Council table is in any position to know if it is the most appropriate to be carrying out this work. In fact, the wording of the document is such that it is designed to achieve the desired outcome with the least possible examination of the appropriateness of the supplier - a potential disaster in waiting, especially when Day has requested such a rapid immediate decision - out of sequence with the Annual Plan - always a grave danger for IT illiterate councillors.

Here are just a few concerns that result from my having discussed it briefly with an expert in the field of implementation:

1. There is no discussion of the applications that the hardware supports.  The document speaks specifically about hardware,  but nothing of the applications that are to be supported.  For example - email is an application that should be in the cloud - no organization these days sets up their own internal email server.  So without a discussion of the software applications, there is no way to assess whether the decision to keep the servers in house makes sense.

2. They are setting this decision up in two stages - should we go to the cloud in some form (IaaS) or stay in house,? and then what should we select as the solution/vendor once that choice is made.?  They spend some time trying to answer the first question - but then lust blow away the second question and give the answer without any real discussion.  A two stage question is the right way to approach it - but it should be thought of as an application by application process.  Email and office productivity apps should probably be in the cloud (MS Office 365 or Google Docs).  Other specialized applications may need to be kept local - but there is no mention of what they are, so there is no way of telling whether they should be kept in house or moved. 

3. Once you have decided on an application by application basis what you want to do - then you have to answer the second question about how to provision the solution in each case.  Here there is an attempt to blind the the reader with the first question and then just blow right by the second.  OK - so we have decided to keep it in we have to answer the question 'how do we provision infrastructure in house, and who do we choose to do it?'  The document mixes and matches the questions and answers to befuddle a non-technical audience.

 4. There is no discussion about the details of what the current infrastructure is.  There is mention of an internal data centre- but what hardware is in it, and what applications it supports is not mentioned.  Without that information no other company could possibly bid for the work.  And there is no way of knowing whether the assertions about the infrastructure are valuable - they talk about disaster recovery but don't talk about what DR is in place today.

5. The concept of 'portability' of the infrastucture - being able to pick it up and move it - that is bogus.  Once you have it in place its not going anywhere.  You do need to have at least 2 locations for a disaster proof solution.

6. The statements about network access to the IaaS services is difficult to assess - i think its probably hogwash. My understanding is that Spark provides, or is about to provide fast (fibre) and reliable internet access from Thames to Auckland where most of the hosting centers are. 

7. This is what is called generic IT services.  Any IT service company can do it, and it should be an absolutely competitive bid.  There is nothing special about this, and there is simply no excuse for avoiding that process.  I it were a custom application built just for TDCC then there may be an argument - but these are generic IT hardware and networking services.

8.  The hardware solution is not a bad one to choose - if you are absolutely sure that you want to keep your applications in house.  But I would be looking for hardware as IaaS services - or simple hosting - before going down this road.  It smells a little like 'a new toy for the IT guys to play with'.  It is rare for a company to keep stuff in house these days - and the talk about it being a 'private cloud' is total BS and marketing from the companies trying to sell you servers to sit in your own data center.  Hardware that you buy and put on your own site is in no way 'cloud'.  

My 'expert' considers the case for this implementation very poorly written,' and clearly seeks to arrive at the the required answers without debate. They have piloted a machine in house given to them free by the vendor, and clearly have not piloted anything in the cloud.  That tells you that the IT firm has sold them the hardware, and they are just trying to justify it at this point.

Perhaps these people should pull their heads out of the 'cloud,' and have their conclusions reviewed, though to be honest, it is probably too late to achieve that outcome. We can look forward to another bureaucratic IT stuff-up if I am not mistaken. You decide!




Article originally appeared on BillBarcBlog (
See website for complete article licensing information.