Making Your Internal Systems ‘High Availability’ Does Not Have To Be An Expensive Dream

When we think ‘disaster recovery,’ we tend to think of grand plans of what to do in the case of the unthinkable happening; a fire, flood or any such incident that renders your office, and more critical, your computer room and equipment unusable. Whereas this is important and we recommend that every organisation have contingency plans in place, the challenge that many of us face is not the big disaster, but the series of mini-disasters we face each day, week and month that collectively represent a major disruption to our businesses.

What I am referring to is unplanned downtime of your business critical systems such as your web server, your order processing system, your email server or you internal intranet and file store. Whereas this might only be just ten minutes here or a couple of hours there, companies are so dependent on such systems to operate that any downtime causes damage. If a potential customer cannot find you because your website is down, they will not tend to come looking again.

Traditional disaster recovery infrastructures such as tape backup, image capture, high-end replication and hardware clustering have failed to keep pace with business requirements for instant recovery at a reasonable cost. We all face tight budgetary constraints and the high cost and complexity of established recovery solutions means that most of us can only afford to protect a fraction of our total server infrastructure – typically only the most business-critical servers. However, this approach leaves the majority of the server network under-insured in the event of downtime or disaster.

What I tend to find is that organisations are protecting nominated business-critical servers such as web-servers and online order processing systems. However, they neither have the budget nor the mandate to protect those servers running internal applications such as email, reporting, and intranet. Ask your sales team what is the most important tool to them and they will say email; ask your customer helpdesk what they rely on and they will say the intranet based knowledge system. Take any of these systems away from these users for only ten minutes and I am sure your telephone will continuously ring.

With everything in life, it is all about timing. It is not so much about how long systems are down but when systems go down. We all know that this only happens at the worse probable time like when that proposal has to be sent to a major new client, or when the sales director has got that key presentation, etc, etc.

Introducing Virtualization
An emerging trend, and what we are recommending more and more, is for organisations to leverage server virtualization to achieve disaster recovery capabilities. Once confined to use primarily in software development, test and server consolidation scenarios, server virtualization and supporting technologies can afford significant cost and performance advantages over conventional recovery options. With the cost of hardware these days, virtualization and high availability is becoming more and more affordable, easily allowing the customer to consolidate and protect workloads which before were not possible, or deemed expensive due to hardware dependencies.

Tape backups can be slow to recover while data replication requires an identically configured standby physical server with duplicate hardware and software, which is a very costly option.

In contrast to these data-centric approaches, virtualisation as a means for consolidated recovery allows organisations to replicate whole workloads (data, application and operating systems) across multiple servers and in the case of a server failure, the workload is picked up by the other server or servers in the virtualised environment. This is simply a case of not putting all of you eggs in one basket; you are effectively distributing your applications across multiple servers and, by default, minimising single points of failure.

Another approach is to put in place a single recovery server or purpose-built device that has an instance of every one of your critical applications set-up in a series of virtual machines operating in a warm stand-by mode; when disaster strikes, you simply bring into service the appropriate virtual machine. This provides significant savings. Rather than duplicating every server you have, you effectively have one server that leverages virtualisation to duplicate multiple systems.

What Options Are Available?
With growing demand for virtualised recovery solutions to provide and extend resiliency and recovery capabilities, solution providers are beginning to bring to market innovative packaged recovery offerings in an effort to help customers streamline implementation, configuration and ongoing administration.

One of the most compelling and cost effective next-generation recovery options is a purpose-built consolidated recovery appliance. This provides a pre-packaged solution of all of the hardware, software, storage and virtualisation components required to implement a virtualised recovery environment to protect server workloads.

Essentially a complete recovery environment in a box, a virtualised recovery appliance, offers many benefits to customers:

• Simplified and accelerated implementation and setup
• Packaged, balanced and right-sized configuration based on the size and criticality of workloads
• Simple, integrated management of the recovery environment as a single entity
• Guaranteed performance
• Fewer IT resources required to manage and administer the solution
• Lower total cost of acquisition and ownership
• Fewer barriers to entry, particularly for SMBs or departmental use within larger enterprises

In the near future, we believe the market will increasingly move toward an appliance-based model for workload protection. A “plug in and protect” appliance model offers a dramatically simplified approach to disaster recovery, especially when compared to conventional disaster recovery infrastructures such as replication and clustering. Purpose-built for protecting server-based consolidated workloads, a pre-packaged, pre-configured and right-sized recovery hardware appliance would reduce or alleviate many of the planning and implementation challenges associated with planning, implementing and maintaining a recovery environment.

To find out more about how virtualisation can help you with your contingency planning, email us at info@b2lateral.net