Datacentre strategy a big downer

When I read the summary of the government's datacentre strategy for the next 15 years, the first thing I wondered was how it could have taken the government months to come up with this document.

The recommendations upon which the whole hoohah was based (Sir Peter Gershon believed the government could save $1 billion over 15 years if it managed its datacentre spend better) were presented in October 2008.

The government took the recommendations to heart and decided to form a datacentre strategy. A cross-agency working group was formed in December 2008, receiving over 50 presentations from industry a few months later. Three specialist reports were commissioned from analyst firms like Gartner.

So that all of this could go on in peace, September 2009 saw an interim datacentre panel put into place to fill the needs of agencies which desperately needed additional services.

I don't know about anyone else, but I was really anticipating the release of this strategy. I mean, it's had more time in development than the current fibre-to-the-home NBN proposal.

So I was pretty disappointed when, over a year after Gershon put forward his recommendations to the government, we get a seven-page document.

Don't get me wrong — I realise that what has been released to the public today is just a "summary" of the overall strategy. Indeed if you can believe the summary, the cross-agency panel looked into deep issues such as datacentre construction, location and life cycle costs; current and future agency infrastructure, management and operation; pros and cons of leasing versus full managed services; datacentre supply and demand; and consolidation and standardisation to reduce costs.

Be that as it may, the seven-page summary basically says this: big agencies need to get their datacentre needs from a panel of approved providers as their current agreements expire (or other major changes occur); small agencies will have to use "aggregated arrangements" organised by the Department of Finance. And, oh, let's not forget: standardisation is good. Also, modern datacentres use less power for cooling than antiquated ones.

Ground-breaking stuff that.

Now I don't have any problem with these recommendations. But I do wonder why it took so long to get to them.