Colin Chaplin is a freelance Infrastructure Architect who helps large organisations transform Microsoft-based infrastructure.
Tailspin Toys is a fictional medium enterprise. It has outsourced its IT function, completed a Windows 7 migration, virtualised its server estate and migrated to Exchange 2010. Does any of this sound familiar? What would the cloud offer? Here are a few thoughts from Tailspin’s IT manager…
How can the cloud be good for our business?
Going forward to the cloud does not need to be a paradigm shift in how we do IT. Nor does it have to negate the benefits and hard work involved in delivering the benefits of email migrations, Windows 7 and virtualisation projects. In fact, we can realise the most business benefit by taking advantage of what we’ve done and gradually steering it towards cloud technologies.
Expertise where it needs to be
Our day-to-day support of infrastructure is a fairly standard model but sometimes we find they are focused on the latest project or issue and have taken their eye off the ball in terms of the basic support and monitoring of the platform, and usually we learn this just when it goes wrong! We don’t really want service credits, just reliable systems with constant attention paid to them. It is far better to have the service delivered by people whose sole reason for existence is to support the service and keep it running, using next-generation availability technologies to allow this to happen.
In theory we should have access to the best experts and industry-leading expertise. When the chips are down however, we rely on our people to be extraordinary together and solve the problem. Going forward to the cloud does not remove the need for our people, but it puts them front-and-centre in the business and enables them to look at delivering value for the business, engaging our stakeholders in ways we’ve never done before. Solutions can be stood up and prototyped in the time it would have taken us just to fill in a new work form. This would be possible because we’re reducing the level of effort required in just keeping the lights on.
More Email, a step away
Office 365, and especially Exchange email, is a natural fit for us. We’ve done the hard work already over the last few years in terms of migrating from an aging, over-capacity and under-performing, slightly chaotic platform to a tiered solution that’s tightly integrated with our Identity and Access Management. Our users have got used to the benefits of a larger mailbox. Moving Exchange to the cloud is less of a leap and more of a progression. Mailbox sizes will be bigger again, and the technology continually made up-to-date. There are compliance features we can only dream of just now; improving, not risking, security. In an a world where a user will have a device capable of receiving email within reach every waking hour, a solution focussed on email servers in rented datacentres and connected via huge pipes seems an anachronism.
End User choice
Gartner are already advising organisations to plan their migration away from Windows 7. The investment to sequence all our applications in App-V and introduce personality transfer via UE-V will really be realised by our ability to rapidly offer Windows 8.1 and 10, not as a massive programme-led deployment, but as part of a choose-your-own-device and natural replacement cycle. These devices will be better managed than ever, where a user on an intermittent mobile connection such as on a train, will not be treated as a third class citizen in terms of device management and monitoring. This extends to devices traditionally managed separately like iPads and smartphones. We’ll use the fact an engineer has happened to have left a machine on whilst at home to push out the latest security patches, software and training materials.
Many services are already not hosted by us at all, yet all of our devices force internet traffic into our network then back out again, even for mobile users. The reason for this is our traditional web-proxy which provides necessary security controls. A cloud based filtering solution means mobile users browsing the web are kept safe and secure, but need never use our network resources in doing so.
Better user data
File servers have not changed much since the 1990s and in some ways are a necessary evil. We have a lot of data on them which is too valuable to lose, but not valuable enough to store and offer in the way we do. Experience tells us that asking for housekeeping does not work. File severs were never designed to work in an internet age, nor work over wide area connections. Users also can’t understand why they might have a PC with a 500GB hard drive, yet we get twitchy from a capacity perspective if they have a 50GB home drive.
There are two ways we can tackle this problem and introduce cloud advantages. For personal storage, users can use the Skydrive Pro feature to have local-hard-disk-speed access to their own data – yet it’s accessible on any device. Where group file storage is still needed there is a way to deliver local-fileserver performance yet leverage cloud-scale cost efficiencies. StorSimple is a technology with an onsite server that effectively acts as a caching engine so recent data is available quickly with the majority of our dormant data transparently hived off to the cloud. The users will not notice any difference (other than no housekeeping requests!)
This is assuming our users need equipment at all. We make heavy use of specialist contract staff. Going forward to the cloud, none of these staff will be provisioned with corporate machines and will instead be expected to supply their own, hooked up to high-density guest WiFi with secure links to our systems and their mobile devices made secure by our mobile device management. They will be up-and-working on their preferred device from day one, not tracking down a spare laptop.
Servers where and when we need them
As part of the virtualisation project, with the emphasis on virtualisation and structuring our application servers into standard sets with known usage patterns, perhaps without realising it we’ve aligned with what is required for a transfer to Azure services, either by transferring technology or even by doing a virtual machine ‘lift and shift’ of servers as-is. Where latency and bandwidth to our sites is a concern, Azure offers expressroute, essentially a private link into the Azure cloud giving higher bandwidth and lower latency.
We can also avoid the problem where we find performance of the system lacking at certain ‘pinch points’, and the remedy is to massively increase the specification of the servers – probably months later when no longer needed. We can flex performance and usage when needed, perhaps even dynamically! There’s virtually no up-front capital investment required, unexpected delays in projects do not see paid-for computing resource lie fallow. We can transform, grow (and shrink) as the technology is available as-fast-as the business can accept it.
All of these changes build on our identity and access management solution. This already ‘talks’ powershell (the language used to control these systems) and is readily able to be extended to further automate and integrate. Perhaps when a developer starts, a personal dev environment is automatically built just for them?
‘The cloud revolution’ is a popular term. However, for Tailspin toys given the relative work we’ve done to our technologies and architecture, moving forward to the cloud is more of a cloud evolution and the necessary next step in doing what’s right to transform and grow.
Do you agree with the thoughts of Tailspin's IT manager? Anything you would add? Let us know in the comments section below, or via @TechNetUK