Re: [bitfolk] VPS failover techniques

I understand that, and it really isn’t a concern for me — I can just pass that cost on. I provide a small number of web-based services to clients; it’s not my bread-and-butter, but simply a small part of the other services that I provide, and my clients usually expect some form of SLA. I don’t mean ‘five nines availability’ or any of that guff, but simply ‘If something happens, how soon can you fix it?’ — it’s about managing expectations. Now, I’m a one-man band, and I do spend time out of the office, so the honest answer is ‘within one week’, which often draws worried looks. We are blessed with good service from BitFolk and problems, when rarely they occur, are usually things like the networking has failed (this hasn’t happened for a very long time now, though) and are usually resolved by rebooting the VPS. What I’m looking for is an automated way to maintain service in the face of the more usual problems — to hold the fort until I can deal with it. I’d like to be able to say to my clients something more comforting, like ‘most issues will be resolved within 48 hours’, which is a lot more palatable. You mention using haproxy internally and I wonder if that is a service that you could extend to us — we provide you with a list of VPSs that provide a particular service and your haproxy manages it for us? As an addition/alternative to that, is it possible to extend the existing monitoring services to be able to automatically shutdown and restart the VPS under certain conditions? Regards, Chris