Getting Started with Snapshots of Entire Dedicated Server

1 post / 0 new
#1 Wed, 04/30/2014 - 22:21

Getting Started with Snapshots of Entire Dedicated Server

We have a dedicated server at GoGrid. CentoOS6 running on a linux box.

The cost for the same CPU/RAM/Bandwith/Disk storage etc. as a cloud instance is exhorbitant. But we had a hard drive problem about 2 weeks ago. Apparently the file system was corrupted somehow. The disk itself I was told by GoGrid support, was diagnosed as OK, but they had to run fsck on that drive over and over and over again, it would take hours to run. The tech guy's keep saying fsck would fail again and again, but they kept running it. I was praying to the IT Gods, sacrificing coconuts and mangos... and 10 hours later the final fsck completed and we got a prompt after re-boot. httpd start; webmin start etc and our sites were up again. It was nerve racking to say the least.

Now I'm seriously looking at back up redundancy options. It is no longer enough to simply mirror the /home folder (which is assigned to DEV2) that contains all the virtual server(s) contents, because we have MySQL databases and a lot of server config. Even it all the data were preserved and the machine died, I would still be faced with rebuilding that box from the ground up (which was our next option if fsck had failed) Sure my data would be intact, but the web sites would have been down for a day or more while I tackled the rebuild (pull the hard drive, plug them in as externals with USB cable, flash the box with a complete new instance of CentoOS 6, install virtualMin, copy the web site data over and blah, blah blah...)

Recently someone visited us here in Hawaii and they said we could get a complete snapshot of our server (I don't mean the virtualServers, but the entire box) on an Amazon Cloud storage plan... for just pennies if we were not actually serving the sites from Amazon.. continue to use our dedicated server (at the good price we get now) but the snapshot would be updated regularly with any changes on the production box... if the production box died, then we could just "spin up" the Amazon cloud back up instance and Viola! our sites would be live again.

I'm just taking baby steps now... and have no background at all in the area... look for some guidance. Is CloudMin my best option for doing the above?