A couple of years ago IBM built an $80 million data centre in New Zealand, to provide hosted and cloud-based services to New Zealand businesses. It was supposed to be state of the art, and have all worst cases figured out.

Until this week … the data centre was down for two days:

“IBM’s $80 million data centre in South Auckland has now been down for more than 30 hours and customers say the outage is having a serious impact on their businesses.

One east Auckland school has been left completely stranded in the same week that it hosts a visit from the Education Review Office (ERO).

IBM said today from Sydney that it had a team of global experts working on the outage as a high priority.“

See:
– Stuff, Fault at Highbrook datacentre (February 19)
– Herald, IBM data centre fault enters second day (February 19)

One of the clients I work with has their infrastructure hosted at this data centre. Not a great way to start the week—what do you *do* when you have no tools to connect with a distributed workforce?

Some thoughts:
– Other cloud-based providers have experienced similar outages, and these have been resolved within a few days. Not great for those on the system when it goes down, but in the scheme of things, it’s a small price. If we think back to public IM systems some years ago, there was even a spate of multi-day downtime for those.
– While you’re in the midst of the downtime, though, it’s pretty horrific. Most feel helpless; apart from calling IBM (or your provider) every 10 minutes, there isn’t anything you can do. With an internal IT system, it seems like you can generate a lot more action in short amount of time.
– My view is that if IBM can get the data centre back up today, this weeks outage will cost them only a few customers. If it drags on for too many more days, or if a similar fault happens again in the coming months, more will move away.

Cloud-based delivery of IT capability offers many benefits. But it must be rock-solid, to mix the metaphors.