Should a problem arise and the service become unavailable, how quickly can the provider get the service back up and running and how much information will have been lost?
Magnus Leask, IT Director, Fast Track,
Whatever the true definition of cloud computing is, the fact remains that as it begins to creep into mainstream IT, a number of key challenges around business continuity and the availability of services, data and applications must first be addressed before adoption becomes widespread.
As the IT director of a sports marketing agency, one responsible for the organisation of key global sporting events, one of my major concerns and objectives surrounds our systems remaining available 24/7. This is particularly important when events are well underway as there is increased demand for remote access to our vital documents by our employees.
Moving business-critical applications and data outside the business to a cloud repository is now a major consideration for CIOs and IT managers, and it is a decision not to be made lightly. There is worry among some organisations around farming out key functions and services to third-party external providers. This is, to some extent, about control and IT being able to manage the process, but there are also concerns about the cloud provider's ability to adhere to service-level agreements (SLAs) and guarantee that the business will be able to access data on-demand.
Should a problem arise and the service become unavailable, how quickly can the provider get the service back up and running and how much information will have been lost? When answering this key question on behalf of Fast Track, the problems arising with cloud solutions amounted to a high level of risk for the business, as we were looking at recovery times of approximately three hours. During a high-profile event, this would simply cause a financial loss of up to £8,000 per hour, which is simply too much to sacrifice.
With this in mind, it is important that companies moving toward cloud options should first ask the question of whether the provider can meet a set of tightly defined requirements around availability. It is also vital that the organisation aligns its business continuity plans with reliability demands of clearly defined SLAs.
Additionally, the business must drill down and focus on what level of continuity is required for individual applications and the business processes that are driven by them. It's clear that some applications are more business critical than others and therefore the uptime of these systems and the need for round-the-clock availability differs greatly. For example, farming out an SAP application -- which is used by the entire sales force -- to an external cloud provider is a much greater risk than a HR system that is used just a few times a week.Getting priorities in order
The key here is prioritising; in other words, defining and evaluating the core applications and which of these can endure downtime without significant repercussions to the business.
Companies should continually ask questions about which applications should be placed in the cloud for DR. These key questions may include: Will a cloud strategy support critical applications such as warehousing, invoicing and SharePoint and allowing the business to continue operations in the event of a failure? And then when an IT outage is over, how do you get out of the cloud and back on-site?
Businesses should also raise concerns about latency and bandwidth, as these will invariably impact the speed at which data and applications can be accessed in the cloud. As an example, it is possible to port older bespoke applications to the cloud, but packages that are not written specifically for the cloud tend to run slowly or may even crash -- which is essentially a Catch-22 and the one thing you wished to avoid in the first place.Moving to cloud
So with all the above in mind, I decided on using business continuity software from Neverfail to protect our most critical assets -- Blackberry Enterprise Server and Microsoft Exchange. Previously, we have had to send our staff home when systems have gone down, and our remote workers have been stranded with no access to e-mail and vital documents.
Previously we have had to send our staff home when systems have gone down, and our remote workers have been stranded with no access to email and vital documents.
Magnus Leask, IT Director, Fast Track,
As you can imagine, this made managing details and making decisions on behalf of our clients extremely challenging. Our primary data centre is housed in its London headquarters, while its disaster recovery site, based on a virtualised infrastructure from VMware, is housed some 30 miles away in Maidenhead. Neverfail monitors the performance and availability of the entire application environment. What we have found the most advantageous is the control we have over our infrastructure which is something you lose when housing applications in the cloud. When Fast Track's infrastructure undergoes maintenance work or modifications, we are able to do this with ease instead of liaising with the cloud provider for a considerable amount of time before any changes can occur.
When the terms backup and availability are being bandied around, businesses need to be very clear what these mean and exactly what they should expect from a business continuity perspective. When you drill down to the complexity of business processes, it is extremely clear that many of the applications which run them are simply not cloud-enabled and to move those outside the business could be an enormous risk.
The cloud clearly has a big role to play in IT as we move into the future, but businesses will always need to step back and ask fundamental questions about which systems they simply cannot afford to lose and which should be kept inside the business at all costs.
For Fast Track, the granular approach to business continuity which allowed us to prioritise key applications and thoroughly protect them with limited lead time proved to be the most lucrative solution financially alongside a paramount level of reliability and control.
Magnus Leask is the IT director at the UK sports agency Fast Track and a contributor to SearchVirtualDataCentre.co.uk.
This was first published in October 2010