This is my biggest worry now that I'm starting my business(not webhosting). All my users have their data on my servers and I'm so scared that something will crash and it will all be gone or the downtime to bring it up will cost them money. Currently I'm backing up all the data nightly to another computer but if my main server goes down I have to quickly load all the programs and data back on the new server and prey it works, probably 2-3 hours downtime. Just curious on what you guys would do if a HD failed or disk controller, CPU, anything....
Well failures are going to happen so plan for them. Kudos for thinking of this up front, many don't.
First try to have redundancy in your server Raid 1 or 5 disk will help if you have a single drive failure.
You seem to have backups already set so thats a good thing. A communication plan is probably one thing that you should have in place so what happens if you happen to have a major outage? Will you be able to get to your customer lists and let them know what is happening?
Be willing to pay for an OS load and emergency support by having an emergency fund available. Placing a call to your server provider or if you host yourself being able to get the hardware replaced quickly is important. Have spares. Of course if you host in your own facility you could consider a disk image on a regular basis so that if you happen to lose hardware, you can use the image to restore. Just a word of caution that this only works if you are going back to same type of hardware and if it wasnt a software problem that took you out to begin with. If its software you could just put the same issue back with an image restore.
Finally, if you are ultra paranoid and if the data is really worth a lot and can afford it you could mirror your data between two servers so that if you lose one, the other is still active but that is fairly complex and not for the faint of heart.
Good luck. You are thinking of the worst case up front. Thats a good thing.
well, we host hotel reservations so if they get lost...that could make many many people upset! :/ I was thinking of having a duplicate server(hot standby) that gets updated every few hours so if one fails at least we can stay up but might loose some data.
The only thing I would add, in case you arent already doing it is regular monitoring and maintenance of the server. Be sure it is secured, updated regularly, firewalled preferably, check the disks every so often with something that checks for bad sectors, S.M.A.R.T status, CPU/Chassis Temperature.
The best disaster recovery plan, is one that is well planned and thought out, and then never used.
...I checkon the servers more than my gir... It pretty much tells me everything and I think I have it well secured. Only one user login, no root, email when ever someone logs in, daily monitor of logs....I'm paraniod!