I’m trying to make a hot spare using my dedicated server, and a VPS which I have access to.
I have installed virtualmin on both servers, my main dedicated server has been up and running for a couple of years now, but there are a lot of times where the server goes down (for a number of reasons).
I used my VPS for backing up (gzip and ssh files over to the vps from the dedicated server), but its now become important to have websites and email up more then it currently is.
I have 2 IP addresses on my dedicated server (websites on both IP addresses) and I have a single IP address for my VPS.
I was thinking heartbeat, played around with it (and killed my dedicated server for the time being till I go and restart it in the morning), but for heartbeat, i need to have the servers on the same subnet, and also have a IP address which isn’t linked to any other websites (this i cannot do)
I was thinking rsync the files between both servers, but how can I control if one goes down, the other server picks up automatically?
Aside from building a cluster sync’d which can sometimes become complex depending on how you go about it, have you considered splitting services into their own nodes?
What we have in our infrastructure is:
dedicated dns, dedicated email, dedicated database, and dedicated web servers.
All servers work together except the email servers which work sort of independently of the rest however is not very difficult to setup.
While initially this setup can depending on your technical skill seem a bit complex, it works great with minimal maintenance after it’s initially setup.
The primary benefits of this design is that each node can be optimized for what role it’s playing, and if the email server goes down it doesn’t affect dns, web, or sql servers. The same goes for most other things as well.
If you’d like to explore this model, please feel free to let me know.
Phone: 604-229-0715 (new)
That would mean adding another VPS to the system, which I don’t really want to do at the moment.
The kinda setup I would like (if its possible) would be two systems, exact same copy of each other, and if one goes down the other takes over. Which would mean a 99+% uptime for all services, instead of if I chose above, one service could go down, if its the database service, it would bring down the web service too (that rely on the database to load).
Mail would be one of the biggest things to keep up really, I have seen the catch all forwarding, but that isn’t really what I would be after either, as I would want the users to be able to receive their mail if one of the servers go down