News

Make no mistake, despite very public emarrassments at Sony and Amazon, cloud computing is here to stay. IT providers need to embrace the paradigm and IT users need to see beyond the hype and plan the best delivery platform for the services their business depends on.

(In fact the cloud paradigm is not so new: BT deliver voice and the BBC deliver TV and radio from “the cloud” to customers who have no involvement with the technology or processes that the services depend on).

Sony left a security loophole which allowed thousands of Playstation users’ details to be revealed. Amazon carried out a network upgrade which went wrong and denied companies access to AWS services for 10 days.

There are two lessons that I would draw from these recent incidents:

One: No one is too big to fail

The first is not to make assumptions about the performance, reliability or security of cloud IT services just because they are delivered by a huge multinational. You would imagine that Microsoft, Google, Sony and Amazon has armies of highly qualified and experienced IT staff, manning vast datacentres providing the finest available IT services. You’d be right, but that didn’t stop all four organisations having falures in systems or security relating to cloud services in the last couple of years. There will undoubtedly be more.

An intelligent approach to using cloud services would be to ensure that you do not place too much reliance on one provider: Amazon client companies who handled the AWS outage best had actually planned for such an eventuality and had a failover plan ready. Some were able to get going again in less 15 minutes.

The same logic that applies to any IT system applies to cloud services: plan for the day when it will fail and don’t assume that it is immune because of a particular brand name.

Two: Reversed economy of scale?

The other lesson is that big providers, despite having virtually limitless resources, do make attractive targets for organised hacking campaigns. It wasn’t heavily publicised at the time but in January 2010, Google discovered that they were being comprehensively hacked by the Chinese. The doors were subsequently bolted and Google even took steps to hack the perpetrators back. However, given that the Chinese government are likely to have every line of Google code up to January 2010, at what point should businesses become confident enough to switch their email to the Google platform?

Oil & gas companies have plenty to worry about on this score since the global energy industry was targeted by a well-organised network hacking campaign in Februuary 2011 – dubbed Night Dragon by McAfee.

Our recommendation is that companies should examine the pros and cons of any IT service, cloud or onsite, from the perspectives of performance, security, service levels and features as well as cost. A knee-jerk reaction aimed at reducing the cost of an expensive IT department could easily backfire if the decision were entered into without some detailed analysis beforehand.

A friend was telling me recently about his company’s general frustration with their IT support provider. (I’d love to quote for a better service but they are unfortunately tied in to a long contract with this provider).

Vicious circle

Despite the fact that the support techs involved are hard working, competent guys with great customer service skills, the client company’s productivity was suffering and frustration levels were rising.

The support people did not seem to value the users’ time and thought nothing of spending several hours on a user’s PC working through an issue. Persistence is often laudable but in this case the user was held back from her work for half a day leading to a knock on effect on productivity throughout the company. This kind of cure often causes more pain than the original problem.

Priorities were decided by a combination of who-shouts-loudest and first-in-first-out. The same faults occurred time and again and lessons learned by one tech are not shared with others.

The support techs’ hard work, dedication and technical skill were simply not enough to deliver the productivity gains expectated from the company’s investment in up to date IT systems. Users and IT provider are caught in a vicious circle where the users’ level of frustration continues to escalate.

Structure

In my opinion things will not improve at this company unless some structure is applied to the support effort, such as the ITIL framework for instance. Some may dismiss ITIL as being over-bureaucratic, designed by civil servants with too much time on their hands or assume it is a job creation scheme created by and for the IT industry.

Our own use of ITIL at ISN has provided us with a toolkit of best practice which deliver real benefits which result in happier users long term. Adopting just a few ITIL best-practice approaches would unglue the wheels and increase productivity for my friend’s company and result in bottom line savings.
Just to pluck a few examples out of the air:

• Having senior engineers dedicated to “problem management” would address the underlying causes of common faults and prevent them recurring
• Agreeing a method with management for “incident prioritisation” would reduce impact on the company’s core activities
• Developing a searchable knowledgebase from information logged regarding problem resolution would shorten fix times

The fact is that most office workers can’t get their job done without reliable IT systems. With the pace of modern working life, any hold up wastes time and money and increases stress levels for end user and IT staff alike. Adoption of ITIL does not have to be all-or-nothing; a company can choose the elements that fit its business requirements and culture.

Once a business has settled on the right server technology and understands the benefits of installing quality hardware that will perform all day and all night for years to come, it is tempting to look for opportunities to save money on secondary storage like the hardware around the chosen backup solution.

Considered from a perspective where everything is running just fine, cheaper backup hardware seems like a great area to trim cost. When the inevitable happens (IT engineers quickly learn that all hardware is going to fail sooner or later) your backup becomes a vital lifeline to get the business back on its feet again quickly.

Let’s consider the common scenario of using a Network Attached Storage (NAS) device as a backup target. This is a great idea since it allows backup to finish more quickly and to take place more frequently than most tape-based solutions would. Crucially, restores can be done far more quickly than from most tape systems.

Home NAS device

Many people would first look at the capacity and then look at the price: gigabytes of storage per pound spent seems a reasonable yardstick. This would make the Western Digital My Book Edition II 2TB NAS look like a fantastic bargain at £190 or so. When you set it up with RAID 1 (mirroring) the storage space drops to 1TB but that might still be enough to back up a small server. To use any disk system without some kind of RAID, mirroring as a minimum, would be crazy. In our experience, such cheap NAS boxes are best for home use, as in this picture, where they are only written to by one machine at a time and aren’t expected to work too hard.

Small Business NAS Device

Scaling things up a bit, we have used higher capacity and higher specification NAS devices as backup targets in environments where cost is a factor. Like the Buffalo Terastation III 6TB NAS shown here, they run a version of Linux, have software RAID 1 or 5, a Gigabit network card and are built to a businesslike standard. Cost is around £900. These work well in general. However when they are written to by multiple servers at the same time, the internal hardware starts to reach its limits, which can result in missed backups. Vendor support can be patchy and is seldom speedy.

Entry Level Corporate NAS

For multiple server environments ISN use a higher grade NAS as a backup target: basically a server running a version of Windows 2008 known as Storage Server. These rack mountable enterprise level NAS devices have hardware RAID controllers, can have multiple network interfaces and a range of other hardware options. Prices start at about £2200 for the HP X1400 2TB model shown here. Dell’s PowerVault NX300 NAS is around the same price. These use standard parts and an operating system which is easy to support worldwide. If you add an onsite warranty pack, you can have an engineer from HP (or Dell) on your doorstep to fix it within 6 hours if it breaks . So much better than trying to revive a cheaper NAS at the time when you need it most.

Value doesn’t come cheap

As in so many situations, you do get what you pay for with IT hardware, just like you do with cars, cameras, smartphones, whatever. The cost of a £3000 NAS depreciated over 3 years is a minimal cost to a business compared to several days of downtime if an inadequate backup system fails to restore your critical data.

This short article only covers one aspect of data backup for businesses. We sell all the items mentioned above and many more besides. Please speak to your ISN account manager or our sales department on 020 7313 8300 to discuss whether your backup solution is robust enough for your business needs.