When we select a Cloud Provider we are effectively relinquishing ownership and control of the infrastructure over to that cloud provider. So how much control do they actually have?
Hardly a week goes by without stories in the press regarding some kind of IT infrastructure failure that has affected its end users customer base. Every Cloud solution relies on infrastructure to support it. So who is really in control of this?
Before I give my opinion of the above I want to see what is happening with the supply of Cloud services.
Other than the big names we already know about, lots of VARs are now attempting, or have completed the move over to Cloud Services. Driven by the demise of their traditional way of business, as end users look to cloud as a new way of sourcing hardware in a more flexible and hassle free way.
So what do these VARs turned Cloud suppliers have to offer?
Well I’m sure they will know their technology of choice inside and out that goes without saying.
But where do they put it?
And how do they manage it?
How are they going to monitor it?
And who does all this?
It’s not something you dip in and out of and it costs a lot of money to set up if you’re doing correctly!
Most won’t have a data centre and the quickest route to market is probably to colo somewhere with a big specialist.
So you have your DC and you know what to put in it, now they had better learn to manage it, or as a another quick route to market they could outsource the Managment to a specialist, or at worst wing it and learn as they go along!
Even the big players can’t always say where the customer data sits in their massive world wide infrastructure. So who’s in control there? And how do you get a company so big to care? If there is a problem it’s likely to effect 100’s if not 1000’s of users.
And now look at control.
In the example above, the VAR has only fulfilled what it did before (i.e. deliver a bit of technology), everything else is out of their control. Their influence on an outsourced Data Center & its management will never be as effective as it would be if they owned and ran it themselves. Go a little deeper and you might find that in BAU even the technology is not in their control. Who delivers the hardware maintenance for instance? The manufacturer? I’m not aware of any VAR that holds sway over any Manufacturer when there is a problem. In fact, I know of one prominent manufacturer that holds all its high value spares for the UK in central Europe. So if their maintenance customers are unfortunate enough to have a hardware fault that impacts the cloud they deliver, what control do they have over the resolution of the issues?
I’m exaggerating to emphasise a point, anyone with any sense would build an environment capable of running even if there were infrastructure issues. But how far do you go? Once you have a failure on your infrastructure you are vulnerable. Your margin for further problems grows thinner. As a customer you need to know your Cloud services can not just withstand an infrastructure failure, but that your provider has measures to resolve it and most importantly has the highest possible control over the situation.
A supplier that outsources even the smallest part of a cloud solution is exposed to situations outside of its control. In a time of crisis that is an additional risk many business’ would not accept.
If the landlord of the DC they are in decides they want to turn it into flats because the money is better, then who’s going to stop him? If the Cloud hardware is down for 2 days because a part has to be flown in to fix it, how do you speed this up?
There are many other things that can go wrong and be out of the end user and the cloud reseller/providers control. And not every VAR will need to outsource.
My point is when we look at a Cloud provider we are relinquishing some control over to them. In doing so we should ensure they don’t just take that control and give it away, spreading it over many other companies. We must question the scenarios that need control and ensure they match the SLA’s required to deliver the service.
At Blue Chip we control everything possible –
Data Centre – we own the lease
Infrastructure – We not only own the hardware in our highly resilient cloud, but we keep spares of every component on site and have our own 24/7 engineers
Monitoring – All support staff are employed by Blue Chip and we own and develop our own monitoring software giving full control.
There are some things you can never fully control, Comms and power for instance that’s why we have 2 x N+1 all round to ensure even if we do have an issue we still have 2 x the capacity we need.
We take control and own it, you should too.