Replacement Hard Drive Just Failed

[JonnyAlpha]

Hi;

I built my NAS about 7 months ago, maybe longer using 3 x WD Green 1TB HDDs, a short while later one of them failed and after much research found that WD Greens were not recommended for NAS systems partially due to there ECO spin down settings. Found a fix to stop them sleeping and replaced the knackered one with the only 1TB HDD that I could get locally from a Euronics XXL , a Toshiba 1TB Intenso.

Anyway my NAS has been working fine and just lately I have been backing up a load of Home Videos.

All of a sudden just know whilst backing up another Home Video the video output froze and I hear a a regular "clunk" pause "clunk" pause "clunk" coming from the system case. So I guess the HDD has failed!!!

Oddly I have never in nearly 30 years of computing ever had a failed disc and now I have had 2 in a year. Strangely though the failed WD HDD is working fine as a TV recording HDD attached to my Panny TV???

When I logged into FreeNAS GUI at first there was no ALERT and when Viewing Discs, they all appeared fine, but then after trying to star a Scrub?? (Not sure why I did this) the alert changed to Yellow with the Alert stating that the Pool was in a Degraded State. When viewing discs the Toshiba had disappeared. Oddly though there were now 2 x View Disc Tabs one with all three drives showing OK and one with only the two WD Drives in a degraded state. I guess this is s GUI glitch.

Anyway I just pulled the receipt out for the Toshiba Drive and it is just over 5 months old so it will be going back!!

Question - As the disc is not showing I cannot take it OFFLINE to remove it so should I just pull it?
And then when I stick the new one in Re-Silver?

Thanks

 

[cyberjock]

First, you can expect the WD Greens to kind of suck unless you do the wdidle tool. You've done that now, but they can and do work well. I have 24 of them in my server and I'd buy them again in a heartbeat.

High failure rates in a system can be the cause of many things. The most common is overheating and crappy power supplies. Use smartctl to check your other disks. If they are above 40C then you can expect them to have a shorter lifespan. There's a google white paper on hard drive life if you want to read more about it.

The light only indicates one thing. It tells you if zpool status reports any problems. When you boot it up it shows as fine because there's no history of problems since bootup.

As for the failed Toshiba drive, if you follow the FreeNAS manual for disk replacement you will be able to do a disk replacement as you are wanting to do.

 

[JonnyAlpha]

Yeh I was happy with the other WD Greens and its odd that the replacement drive appeared to fail also? I switched of my NAS removed the HDD and plugged it into a USB Ext HDD enclosure to run some tests. Fired it up and got the normal spin up and access noises and NO clunking. Windows XP picked up the drive fine (clearly cannot see any data as its a FreeNAS Disc. Ran HD Tune and the screen shot is attached - seems fine.

Could it be a dodgy Motherboard or PSU?

I do have another Mobo (Not as good but has 4 x SATA ports) knocking around

I'll have a poke around and put the Toshiba back in and run the smartctl test on the other drives.

 

[cyberjock]

Well, there's no real solid info in this thread so far, so any cause is a possibility. Heck, it could even be user error.

I would definitely look at the smartctl stuff for your drives. If you don't know how to read the info and understand it just post it here.

 

[JonnyAlpha]

Sorry forgot to add the attachment.

 

[cyberjock]

I don't care about benchmarks. Those don't tell you if something is wrong with certainty. SMART, on the other hand, was designed for just such a purpose.

 

[JonnyAlpha]

OK; sorry for being a NOOB but how do I view (and copy and paste) the output of smartctl, i understand the options but when I run it from the GUI Shell I cannot expand the Shell window to view all of the info??

Anyway I swapped out the SATA cable for a different one, bunged the faulty HDD back in the NAS box and its back up and running. Just want to run some smart tests to see if anything is up?

 

[cyberjock]

smartctl -a /dev/XXX > /folder/file

 

[JonnyAlpha]

Managed to run a smartctl but noticed when I logged into my FeeNAS the alert went yellow stating that one of the devices was in an unknown state. Viewing discs showed one of them had CHECKSUM errors.

Anyway all devices were still visible and I managed to capture a smartctl output, ada0 - The drive I think has the problems is the file called smart test .

Three log files for all of my drives are attached.

Any feedback would be appreciated?

 

[cyberjock]

The Toshiba disk(ada0 I believe) is having problems. The reallocation count is 219.

The other 2 disks are okay. I'd replace the Toshiba disk. Not sure if its possible to do an RMA, but I wouldn't trust that disk with anything at this point.

 

[JonnyAlpha]

So just so I know for the future the RAW_Value is the SMART readout from the drive, so in the case of the Toshiba the reallocated_Sector_Ct = 59 of 100, does that means its used 59 of its spare sectors?

What is the Reallocated_Event_Count

The HDD is less than a year old and I have the receipt, I'll print out the SMART tests and take it to the shop???

May I make the presumption that the only thing that could cause a drive to re-allocate a sector is a bad drive? What I mean could this be caused by a faulty SATA cable / motherboard / PSU??

Thanks for your help as always