another dead hard drive....PARITY NOT VALID: DISK_DSBL

[mp328]

wtf, in less then 10 days since i started using unraid two hard drives died on me.

 

last week, the first day i installed unraid the WD 1.5TB green drive dies, so after this was one was tossed out, i was able to create an array for the first time and fill it up with smaller drives. today my parity seagate 1.5TB dies

 

wtf is going on.

 

last night i was having some issues, that the box wasnt booting up after i tried rebooting (installed some scripts), so after 4-5 tries it finally booted up and showed up with parity errors, then it started a parity check. when i checked the smart output on unmenu, it showed around 1230 reallocated sectors or something like that

 

then when i woke up this morning, the drive was disabled. i rebooted the server and the drive showed up as green again. tried another parity check and couple hours later it crapped out again.

 

i think for now i'm just gonna shut it off and not use it until i get the 2 replacement drives in, before i lose the remaining data that is now there without parity

 

the syslog attached is huge, but honestly all jiberish to me lol. the parity disk is SDB btw.

 

i never expected to have to get all new disks to be able to use unraid....

syslog-2010-03-19.zip

[terrastrife]

unraid relies on perfect disks. its quite possible to be happy running windows on a hdd that is 90%ok and you would never notice.

 

is your psu up to the task? unstable power can cause drives to be intermittant.

[SSD]

I think it is more likely that you have a drive connection issue and not multiple drive failures.

 

The most common issue by far in new arrays and apparent drive failures is bad or lose SATA cables causing garbling of messaging between the drive and the computer.  Sometimes just unplugging and replugging cables is enough to fix things, sometimes you need to buy new cables.  Locking SATA cables are very helpful with some ports / disks.

 

To tell if a drive is failing you need to run SMART reports.  Just because unRAID kicks it out of the array does not mean the disk has failed.

[mjstumpf]

It wouldn't surprise me a bit if these were new drives.  Failure rates are at 25% for most manufacturers at the moment.

 

Some people have reported that a beefier, more stable (how the hell do you measure that without an oscilloscope?) power supply stopped constant seagate dropouts.

 

 

[mp328]

i dont know, im using an antec 450W truepower supply in the unraid box, im using locking SATA cables as well on 2 out of 4 drives. the parity drive that failed being one of them that had locking sata cable.

 

im gonna pull the drive out from the unraid box and try running seagate tools on it on a separate computer, see what that thing says.

 

other then that i'm off to the store to pick up another 1.5TB. it's been 7 days and WD still hasnt shipped out my first replacement.

[pfp]

unraid relies on perfect disks. its quite possible to be happy running windows on a hdd that is 90%ok and you would never notice.

 

 

I tend to believe this as I also had two disks fail (out of 11) when I moved to unRAID.  Luckilly I had already copied all my data off them before they really went int othe toilet.  I hadn't seen any issues with the disks on Windows.

[mp328]

well after running seagate tools in dos, it fails on the short test, and the long test completes with about 50 errors where it then tries to recover the bad sector and then marks the drive as passed, but if i rerun the test right after again, same thing happens. after booting into windows the drive doesnt even show up.

 

so fook it, i just send in another rma request and went and bought another 1.5TB green drive for parity in the mean time

[bcbgboy13]

What is your configuration and how old is your Antec 450W "true" power supply as two different disks from different vendors failing in a row is a not a good sign.

Did you run the memtest?

 

Can you measure your 5V DC and 12V DC outputs with multimeter when your computer is on or at least provide the voltages from your BIOS - "PC Health monitor section"

 

I do recall from memory some older Antec PSUs providing close to 4.75V rather than 5V or 5.25V right from the factory so it definitely can mess up with the HD electronics.

 

Good luck.