How serious is this...

M

memilanuk

n00b
Joined
Oct 7, 2010
Messages
20
I have an old PC running FreeNAS from a usb stick, with three 500GB hdds in a ZFS RAID-Z pool. For the most part, it sits there on my LAN, quiet and dependable. Wish I had more storage, but for now it works well enough.

Today I was doing something I should have done originally when I set it up: configuring email status reports. While doing so, I took a look at the SMART log, and found some entries that gave me some concern:


Code: 
Jan  9 04:59:01 freenas smartd[2046]: Device: /dev/ad6, SMART Prefailure Attribute: 1 Raw_Read_Error_Rate changed from 110 to 109
Jan  9 04:59:00 freenas smartd[2046]: Device: /dev/ad4, SMART Usage Attribute: 194 Temperature_Celsius changed from 110 to 111
Jan  9 04:29:01 freenas smartd[2046]: Device: /dev/ad6, SMART Usage Attribute: 195 Hardware_ECC_Recovered changed from 64 to 66

Basically, the log was full of these. Mostly the Usage Attribute: Temperature_Celsius, but all the ominous-sounding 'Prefailure Attribute' warnings were for the same drive - a Seagate 7200.10 500GB SATA drive.

What else should I be looking for, and how imminent is failure ;) ?

Monte
 
Last edited:
I can tell you I started getting the same error about 2 weeks ago and my system started freezing a few days ago. It's still running ok but if I shut it down sometimes I get a squeal when the drive first spins back up.

Let's just say I've been backing up pictures this weekend and my new drives are supposed to be here Monday morning.

I thought I could ignore it for 6 months or so and then worryabout it but the system is telling me otherwise. I'm pretty sure I have a bearing going out on my 12 month old WD Blck 1tb...
 
Temperature_Celsius changed from 110 to 111
Click to expand...

That seems really bad. My WD 640GB Black in my overclocked i7, with F@H and dnetc keeping the CPU and both GPUs at 100%, is showing a max of 40°C. If it truly is over 100°C, I'm surprised the disk is still working.
 
This is normal; smartd just tells you about changes in the SMART data as reported by your HDDs. Those properties are 'pre-fail' properties; meaning that if that property exceeds the threshold (lower than threshold) then it would cause an imminent failure warning.

That means that everything is alright as far as these logs show. Real dangerous SMART properties to look out for, are:
- Current Pending Sector (should always be zero; non-zero means you have an ACTIVE bad sector which your OS cannot read should it want to; bad bad bad!)
- UDMA CRC Error Count (cabling errors)
- Reallocated Sector Count (passive bad sectors; no problem if this exceeds zero; but if it keeps growing this may be a sign of failure as well.
 
Invisibill, temperature raw values are always accompanied by an offset. A value of 110 probably means that 30 degrees C is for 100, so 110 would be 30 + 10 = 40 degrees C. It depends on the drive.
 
You must log in or register to reply here.
Back
Top