Thread: How Often Do Server Hard Drives Fail

In the past few years, I have had 3 hard drives failure: two involving a primary drive, and one a backup drive. This equates to around one drive failure per year. Until recently, I ran 3 servers, each with 2 drives. Therefore the failure rate per year is around 15%.

I wonder whether this is normal or not? The servers involved are now quite old i.e. over 2 years. But it is entirely possible that the drives were not new as datacentres can re-use drives.

gave me a picture of ur server status
to trying help u

Thanks, I am not sure what you mean.

The server is a dual xeon 3.0. with 2GB RAM and Dual 250GB drives.

3 Drives in 3 Years isn't obscenely abnormal.
It might be a little on the high side, but not so much so that alarm bells should be going off.
Not after 3 drives.

If that rate continues, though, I might be concerned.

sorry

i want to see your server status

in whm
Server Status
>>>>>
Server Status

and take a picture to the screen

i want to see the status

because i am trying to help u with it

That will only show you how full the drives are and other information such as server load. It will in no way help with determining why hard drives fail.

I had a rash of HD failures a few months ago- I think it was just an age thing as I ended up replaceing 2 of the drives in my RAID 3 months apart from one another- and then I replaced the RAID array all together due to a lack of faith

Originally Posted by benjaminp

That will only show you how full the drives are and other information such as server load. It will in no way help with determining why hard drives fail.

I have to agree on this. Although I can post some figures on the current loads and usage. This only reflects the current situation with the replaced disk. It does not reflect that happened before the disk was replaced. Besides another server was also involved.


Server Load 3.31 (4 cpus)
Memory Used 36.1 %
Swap Used 2.69 %
Disk /dev/hdb1 (/backup) DMA+ 61 %
Disk /dev/sda6 (/tmp) 3 %
Disk /dev/sda7 (/) 12 %
Disk /dev/sda8 (/home) 36 %
Disk /dev/sda1 (/boot) 17 %
Disk /dev/sda2 (/var) 98 %
Disk /dev/sda3 (/usr) 33 %

Unfortunately, the datacenter made an error with the partition and had given me a drive with very little Disk /dev/sda2 (/var). This has led to several database driven sites shut down in the past week. It seems that it will take some time before they can correct the problem for me.

Originally Posted by zxpro168

In the past few years, I have had 3 hard drives failure: two involving a primary drive, and one a backup drive. This equates to around one drive failure per year. Until recently, I ran 3 servers, each with 2 drives. Therefore the failure rate per year is around 15%.

I wonder whether this is normal or not? The servers involved are now quite old i.e. over 2 years. But it is entirely possible that the drives were not new as datacentres can re-use drives.

It depends on the kind of disks. SCSI disks and 'Enterprise' grade Seagate disks are warrantied for 5 years, they don't break much. Regular SATA disks are only warrantied for 2 or 3 years and should not really be used on production servers. You get what you pay for.

I'm waving at Zap on this one..

when it comes to Hard Drives, i tend to re-use them as much as i can to save costs.

Iv an old Dell which has been running on the same 'ol hd for the last 4 or so years..

Though it does depend on how much pressure you put on the drive, and how 'used' it is. But drives these days are built to last for years-and-years

If you feel the drives are unstable.. Id suggest having them checked out by a geek

lol depends we have 40 servers and never got this problem

Originally Posted by prolimehost

lol depends we have 40 servers and never got this problem

You have 40 servers and you've never had a hard drive fail? You have to be the luckiest person in the business then.

Seems a bit high to me. I would not be to worried as others have said tho.

Man you are really unlucky with your hard drives! I had 11 servers running at one point and only 2 hard drive failures. Most of these servers ran for 4-5 years before I got new ones...so Im betting the hd could have lasted another few years.

You dedicated hosting company could be using junky hard drives which would explain the failure rate.

It depends on how frequently your data keeps changing! buy a branded HDD for ur server! could last a year or teo longer imho!

I'd want to look at there being other problems with how the servers are being run if they are going that frequently. I had the same 1GB server HDD running for six years in the early days, and now tend to upgrade before they fail on me. Once a year is rather unusual.

This is an old post I just came across. I wonder if the user is still having those problems. I have been using the same hard drive in my server for 2 years without any problems. Maybe people with drives that fail easily have to look into what programs are writing to the disk a lot. Maybe there are huge log files or cache directories constantly being written which causes stress on the hard drive. Another possibility are cheap drives that fail easily.

Depends how they are treated and the environment that they are used.

I currently have 8 servers working for more than 1.3 years and no hard drive failures yet.

- Daniel

Some people I know seem to have lots of bad luck with server hard drives failing. In the 5 years that I have had dealings with servers I have not know a drive to fail, this might just be good luck though.

When I did hosting, I had 2 servers for a little over 2 years, and LUCKILY I never had a HD fail, but a friend of mine went through 3 of them in 1 server alone.. I guess it all depends on the variables..

The real issue for me is whether my datacenter is using used hard drives for my servers. I suppose I will never find out. So far, my servers have been running well since then.