Almost in the first sentence of this post about High Availability, we see that a highly available system is highly complex. This is where acquiring the right equipment for the right job is the right decision. Reading on we see the usual situation of an active/passive architecture and you know what is coming …

This always causes interruption to services and downtime – if the active node fails, we have to detect that failure and then instigate a restart of those services on some spare resource – in this case another server.

I am not saying there is anything wrong in doing this and it depends entirely on what service level you are trying to provide, but there is a much simpler way …

The right tool for the job could well be a fault tolerant server – ultra reliable to give you the uptime assurance that you need.