I live about 1,000 km from my office which means that I do a lot of travelling by air. This also means that I get to fly through a lot of cloud. One thing I can tell you from all the clouds that I have been through is that I have not seen any disk drives for computer hardware up there.
Of course that is not the case, and what a stupid assertion, but I make a point because people seem to act as though cloud systems really are in the cloud. What I mean by this is that no consideration is given to where exactly the server and storage resides and how secure it is. The wonderful thing about a cloud system is that it takes about half an hour to sign up to and within an hour you have access to a fully working system through a web browser from wherever you are.
We have been managing hosted servers on behalf of our clients for over five years now and the performance and reliability have been very good. Until last month that is, when a client of ours whose dedicated server is with one of the major global service providers was without access to their server for 32 hours.
When the support call was logged we received notification that a “node” had gone down and as it was a public holiday in Germany there was no technician to put it right. We received the standard railway apology, “we apologise for any inconvenience caused” and that was all. After we wrote a stern letter to the management of the company we received a further apology and a pledge to refund one monthly payment.
The monthly payments in question was less than £400. The real cost of this problem was equivalent to giving all the accounting staff unpaid leave for a day and a half and the cost of slipping a gear in all business processes, collecting payments, paying suppliers and providing the month-end accounts on time. There was also the uncertainty bordering on panic about whether emergency measures needed to be taken to start rebuilding another server as there was no confirmation about when or whether this problem would be resolved. It was news to us that the server actually happened to be in Germany we thought it was in the US!
We do provide a backup service for our clients but inevitably this is through another cloud-based system!
What came home to me about this particular system failure was that the resolution to the problem was completely out of our client’s hands and out of our hands. That’s a dangerous place to be in the midst of a crisis.
I have no doubt that web-based everything is the future (with one major exception – sorry see my next post) but I would recommend to all businesses planning on implementing a cloud ERP system or a dedicated server that they consider what they would do in the case of the blackout. I don’t think it’s good enough to rely on the “always up” assertion from the cloud provider.
Last year a major cloud-based retail system for small businesses in the US went down twice. Apparently it was a power supply fault in a particular district. If you’re a small business with a few point-of-sale terminals in your shop you would not have been able to properly record your sales at that time. That’s pretty serious.