Infrastructure
Our physical system is based on the Amazon EC2 infrastructure. This allows us to focus on providing an excellent cloud hosting service while allowing the experts at Amazon to take care of running the data centers. You may still think of Amazon as only an e-commerce store but for the last 5 years, and by utilising their experience from running their giant e-commerce store, they have been a top Infrastructure as a Service (IaaS) solutions provider.
We are currently only operating in the Amazon US East (Northern Virginia) data center but will be opening up in their other data centers soon. Once this happens you will be able to choose where you want to host on a per-site basis. The other data centers operated by Amazon include US West (Northern California), EU (Ireland), Asia Pacific (Singapore), and Asia Pacific (Tokyo).
The challenge
Designing reliable hosting solutions, even on the AWS cloud, is still a tremendous challenge. A hosting machine on the cloud will tend to fail far more often than a standard dedicated server and any solution, based on the cloud, needs to cater for that. Since we are aiming towards affordable hosting, running fail-over machines for each of our sites would not prove too cost effective. Instead we have opted for a proactive responsive approach. At the first sign of trouble with a machine we will instantly start up a replacement and have all your sites moved over within a matter of minutes.
Reliability and scalability
The scalability of our system is also achieved by using a similar proactive approach. We store each site on a separate virtual disk (see data storage under technical details) and can easily switch the disk to a new machine in minutes. If you wish to upgrade your site from the one package to another, we simply setup your account on a new machine, disconnect your virtual drive from your old instance and connect it to the new one.
Our databases run on separate servers using the Amazon Relational Database Service (with Multi-AZ). This is a reliable database solution that has an automatic fail-over mechanism to a secondary database in a separate availability zone. That means if the machine running your database stops functioning your website will be immediately and automatically switched to a secondary machine running an up to date copy of your data.