- Jun 22, 2012
- 1
- 0
- 10,510
I'm looking for some advice on the possible causes of unreadable sectors on a new hard drive.
I have an HP EX475 Mediasmart Server that is about five years old. It is basically a home server containing three sata hard drives with software raid, one of drive is the system/boot drive, and it runs a variant of Win 2003.
About four months ago I started to get a lot of read and write errors on the system drive. Since the drives were nearly five years old I decided to replace them all. I have good backups. I bought three new Samsung EcoGreen F3 512GB drives (chosen for low noise) and re-built the system without any problems.
Now, four months later, I'm getting unreadable file record segments on the system drive again. The affected clusters are in three groups of adjacent clusters, and the number of clusters affected is increasing. I have not seen any write errors yet. I've checked the other drives with chkdsk and Samsung's test utility (seatools) and they are okay.
I'm wondering what might be causing this and what to do about it. Possibilities that I can think of:
1. I just got a faulty drive and it happened to go in the system slot. Possible, but seems unlikely. I could buy another new drive and see if that fails, but thats more time and money and wouldn't necessarily point to the cause. The system's temperature monitoring is showing normal, so I don't think I'm cooking any of the drives.
2. Power supply. The system has power monitoring and all voltages appear stable within about 0.1V of nominal. I could buy a new one for UK£65 which is fairly expensive and might not fix it.
3. Drive controller/interface. I don't know much about failure modes for these. The drives plug into a backplane, so I can't swap any cables.
I realise that this a fairly vague, but I'm looking for advice on how to approach finding the root cause of this. I'm a developer rather than an IT/hardware specialist, and my hardware knowledge is limited.
Thanks for reading this!
Andy
I have an HP EX475 Mediasmart Server that is about five years old. It is basically a home server containing three sata hard drives with software raid, one of drive is the system/boot drive, and it runs a variant of Win 2003.
About four months ago I started to get a lot of read and write errors on the system drive. Since the drives were nearly five years old I decided to replace them all. I have good backups. I bought three new Samsung EcoGreen F3 512GB drives (chosen for low noise) and re-built the system without any problems.
Now, four months later, I'm getting unreadable file record segments on the system drive again. The affected clusters are in three groups of adjacent clusters, and the number of clusters affected is increasing. I have not seen any write errors yet. I've checked the other drives with chkdsk and Samsung's test utility (seatools) and they are okay.
I'm wondering what might be causing this and what to do about it. Possibilities that I can think of:
1. I just got a faulty drive and it happened to go in the system slot. Possible, but seems unlikely. I could buy another new drive and see if that fails, but thats more time and money and wouldn't necessarily point to the cause. The system's temperature monitoring is showing normal, so I don't think I'm cooking any of the drives.
2. Power supply. The system has power monitoring and all voltages appear stable within about 0.1V of nominal. I could buy a new one for UK£65 which is fairly expensive and might not fix it.
3. Drive controller/interface. I don't know much about failure modes for these. The drives plug into a backplane, so I can't swap any cables.
I realise that this a fairly vague, but I'm looking for advice on how to approach finding the root cause of this. I'm a developer rather than an IT/hardware specialist, and my hardware knowledge is limited.
Thanks for reading this!
Andy
Facebook
Twitter
Instagram