I’m not new to the forums, but I believe this is the first time I’ve posted something here.
I have a probably not to uncommon situation at the moment that I’d like to get suggestions on in general from thsoe of you who might have delbt with such things.
At the moment, I only have a single server (as I don’t have the large budget that multiple servers of any type VPS or otherwise would require).
Because I only have one, and because I’m not going to promiss people downtime when that’s simply something I cannot keep 100 percent of the time anyhow, and because maintenance is something I do regularly, I face a small issue.
That bbeing, that whenever I do take the server down for an extended period of maintenance, like checking the filesystems offline, etc and so on, no one can access my sites, etc, mail doesn’t work at that point, etc.
So I’m wondering, what are some things you folks have done to keep that from being a problem if you only pay for or own a single server, and don’t have a budget to justify paying for multiple separate servers?
Just curious as these suggestions (depending on what they are) might actually be useful.
Today is a good example for instance.
Because I’ve taken the master server (and for now only server) down that hosts my stuff, while I do some stuff that simply needs to be done offline entirely, my main websites etc don’t work, and mail directed at my email server won’t work, either, not to mention DNS failing to resolve.
Not sure if my answer is useful to you.
The servers I have for webhosting are for just that, webhosting and I dont take them offline, unless for a reboot after kernel update.
For everything else I would need a box for, I virtualize and for development I do that mainly on my workstation at home which is also virtualized.
Now to take a box offline for a long time, this would mean to have another box with same contents and a heartbeat server so you can switch to another box. But if that is not in the budget I would say, don’t take the box offline…
Mail can be done through gmail or if its not too extensive, have a small emailserver at home. This can be a good PC/workstation or a small NAS.
Well, budget is a big factor there – there’s many ways to accomplish that, which one is used comes down to a number of factors, a large part of which is budget.
It’s actually not uncommon to only have one server available – in fact, many folks run their Virtualmin servers on a single VPS.
In regards to maintenance, it’s not uncommon at all (even when dealing with multiple servers) to have a system be unavailable during a maintenance window. The issue there is largely just a matter of giving customers notice, as well as doing the maintenance at night.
As far as disaster scenario’s go – I think the main key to consider, is “how can I get the system back online as fast as possible”.
If your main server were to die, and you don’t have a second server – that means you’d need to rebuild your server, install Virtualmin, and then populate it with your data.
If you have a place where backups are stored – you could store your Virtualmin backups there, as well as any config file changes.
And then, downtime would be the time it takes to reinstall your OS, install Virtualmin, and then restore your Virtualmin backups and config file changes.
Now, if you’re responsible for your server’s own hardware, you’d want to make sure you have spare parts on-hand, or else that problem can go from hours to days while you wait for spare parts to be shipped to you