Will a third Hard drive ensure better server uptime?

[radical]

i have noticed that when a server primary drive develops problems, the Datacenter takes nearly 24 hours to replace the drive with the OS, and backups take another couple of hours. the usual response would be that OS reloads take 24-48 hours and we are in the queue. so to avoid this, i was wondering if i opt to take a third drive from the DC, and have them install OS and Cpanel in it. so if and when the primary develops problems, i just ask the DC to plug in the third drive, and start restoring backups from the existing secondary drive which has the backups.

can some knowledgeable soul here, tell me if this is a workable solution, and if there are any loopholes that need to be covered...

thanks

 

[webignition]

A better option may be to have two drives in a raid configuration, with the third (currently second) drive used to store backups.

With two drives in a raid configuration, both will have identical content i.e. one will be an exact mirror of the other and will be seen by the OS as being one drive. As long as both drives don't fail at once you should have no problems.

If one drive fails and is replaced, it is commonly the case that the raid array is restored before things continue as normal i.e. you'd have to wait X minutes/hours whilst the new drive is mirrorised.

You'd have to query your DC with regards to what raid options they can offer you. Generally speaking, the more drives in a raid array the lower the chance of losing all of your data.

Raid arrays work best with identical drives i.e. same model number. Furthermore, you should expect an operational lifetime of up to 5 years for a hard drive.

Consequently if you start out with a brand new raid array there's always the chance that, with both drives being of equal age and having been used an equal amount, they may start to fail at the same time once they approach the end of their operational life.

With a two-drive raid array, I'd recommend replacing one every two years.The first replacement will be for a drive that is only 2 years old, but from then on you'd be replacing 4 year old drives once every 2 years which should reduce the chance of sudden failure due to 'old age'.

 

[radical]

chances of Software corruption in RAID1

thank you for your reply. i had earlier used an IDESATA RAID1 config provided by my DC. but within 3 months of getting my server, one of the hard drives failed. the dc staff told me that the hard drive failure had caused some corruption of data and since both were mirrored, both the drives could not boot up. Since we are located half way across the globe, i really could not believe if what the DC staff told me was true or not. my question is what are the chances of having this kind of a problem, especially if you are running RH with Cpanel. My bother is that RAID costs a lot, and if it does not help at the time required, then it is money down the drain!