Disaster recovery

We believe that planning for failure is a crucial component to running an effective network service. When it comes to your own servers, you need full, automated remote control and that’s what Bytemark has given its customers from the start. We promise that in the event of trouble, you’ll never need to wait for an engineer to visit your server to reboot, reinstall from scratch or fix a broken network configuration.

How is this possible? You get to use our excellent dedicated host shell which provides these crucial facilities:

Serial console

On a Linux server, the serial console is very similar to have a screen and keyboard available—you can fix things that would otherwise require a physical presence at the server including network configuration, boot problems and supervising an operating system installation.

This means you are never stuck for control

Network boot

All our servers are configured to try to boot themselves via our network server before they look at their discs. Our network server makes sure that a menu of alternative boot options is presented—if no response is received from the user, the normal boot continues. However with the administrator present, they can choose to run a normal OS installations process, a special “rescue” mode for helping them out of configuration scrapes, or RAM diagnostic program.

Power cycle

We use APC switched PDUs which allow full remote control of the power to your server, to allow you to force a reboot when all else fails.