Last Updated on Monday, 28 December 2009 11:43 Friday, 07 August 2009 09:16

Q: I need a better disaster recovery plan for our servers, but isn't a SAN too expensive compared to a NAS?

A: Since SAN (Storage Area Network) and NAS (Network Attached Storage) are often used interchangeably, let's look at the difference between them.

A NAS is an entry-level fileserver using a limited number of common hard disks and common network protocols. It is simpler and often more reliable than a traditional fileserver, but more prone to attacks. The trade-offs of using a NAS are that it is difficult to protect it and its files from hackers, virus infections, and other network threats, so it is only recommended for very small startup businesses or home offices.

A SAN is a network of hard disks limited in size and speed only by the disks and network used. It can be configured with a wide range of components from low-cost common hard disks to more expensive faster ones and from normal Ethernet to the fastest gigabit and fiber networks. It utilizes modern network protocols used by servers and the latest computers so they access the SAN as if it were local storage, but at speeds typically much faster.

In a business with more than one server where planning for future expandability is a must, it is worth looking at a SAN. In the last year, entry-level SAN products have begun to appear at a fraction of the traditional price. Using open source software and common server components, a SAN can even be built for much less than a traditional server, or even using existing servers and components. For instance, a SAN could be part of a move to virtual servers when buying a small number of new servers and using old servers as backups or SAN components.

So, the benefit of using a NAS or SAN in a disaster recovery plan is that they separate server storage from the server itself, protecting the data from downtime if a server fails. Two or more servers could use a NAS or SAN for their data storage, with some servers acting as backup for others. The trade-offs of using a NAS mentioned above negate the benefit of using it for disaster recovery unless it is completely separated from the main network to protect it. Using a SAN to benefit disaster recovery is recommended unless all options to use a SAN are unacceptable.

In the past, the biggest obstacle to using a SAN was the cost, which was typically two to six times that of a single server. Now that low cost redundant disk arrays are common and reliable, SAN hardware can be built at a much lower cost. The controlling software is also now available from at least two open source projects for free. So, now a new SAN with around three terabytes (3,000 gigabytes) of storage can be built for around $1,000, about half the cost of a typical entry-level small business server.

Using a new generation SAN will benefit your disaster recovery plan while also laying the groundwork to start moving existing physical servers to virtual infrastructure now or in the future. Virtual servers bring cost savings by better utilizing existing server resources, reducing future server expenditures, and the higher availability further benefits your disaster recovery plan.

A typical virtual infrastructure server setup with two servers (one for backup) and a SAN containing 1,500 gigabytes of storage can host the equivalent of up to thirty-two physical servers with their typical resource requirements. The cost would be around $2,300. Aside from the obvious future cost savings in avoiding new server costs, there are immediate savings from reducing power and cooling costs by consolidating servers if you already have more than two servers.

So, there are ways to use a new generation SAN to not only improve your disaster recovery plan, but in a way that pays for those improvements in a matter of months. Contact us for a complete analysis of the benefits to your business.