Scalability, is it important?

What do you do if the hits to your server expand at a rate comparable to new users comming on to the Internet, or your new web application consumes more resources?

You rely on scalability, wether you scale-up or scale-out depends on the type of service that you provide, but with a real server you know you have all the options available to you.

Where possible, our usual advice is to scale-out and load balance. This can be accomplished with either a low-tech apporach such as round-robin DNS, by using software or specialised hardware. We use several types of hardware load balancer to ensure that your customers never receive an error message. Errors caused by a server hardware failure or software faults are recovered, the requests transparently re-directed to an available server; all this without sending an error to the end user. What if the load balancer fails? We always use them in pairs to provide redundancy.

This gives you the best possible flexibility, price performance ratio & and failure protection, while we deal with the increased administration of multiple nodes and load balancing resources.

Of course with services like MS SQL Server, which cannot easily handle load balancing transactions, the way to scale is by adding further CPUs and add redundancy by failing over to a clustered standby server.

Whichever is best for you, be assured that Arrowhead can offer the best hardware with the correct specification to suit your application.