Hi Guys,

 

I'm in need of some help.

 

I build my unraid server 1.5 years ago now and everything has always worked flawlessly, never really had a problem more then i just had to learn how to use unraid. I have 2 2tb WD green drives and a samsung 2tb drive in my array with a 1 3tb WD green parity drive.

 

The 3 TB parity drive is only 6 months, but in the last few days i noticed that it had a red ball next to it in the web interface with a few hundred in the error column, I also had a few hundred in the error column for the Samsung drive but its not red and i can seem to still access my array.

 

however i'm really worried that i might have problems with the data on my array.

 

I have tryed to re-sync the parity a few times now and it seems to go well each time until it gets to the end where it fails. What % im not sure because i haven't seen the last 10% of the parity build, but before that there are no errors and everything seems fine.

 

I have tryed changing sata ports and changing sata cables on the parity drive and the samsung drive.

 

I have run a short smart test on the 2 drives. they both passed, but the parity drive had a raw value of 99 for the Current_Pending_Sector. I read this is bad but what do it mean I should do?

 

I really have no idea what to do, so any help would be really appreciated. I'm really don't like having a unprotected array its kinda scary!

 

My syslog got reset this morning when i turned off the array and restarted it so i'm doing another parity sync to get a sys log after the sync fails and after i get those error numbers i mentioned, and ill post that as soon as its done. However I have attached the short smart test results for parity drive and the one in the array that gets numbers in the errors column after an attempt to resync parity.

 

any advice i would greatly appreciate

 

edit: im trying to run the parity syn again but it running at 300~ kb/sec normally its ~115 mb /sec. any idea what this could be??

Smart_Test_-_Parity.txt

Smart_Test_-_Array_Drive_with_errors.txt

The 3 TB parity drive is only 6 months, but in the last few days i noticed that it had a red ball next to it in the web interface with a few hundred in the error column, I also had a few hundred in the error column for the Samsung drive but its not red and i can seem to still access my array.

Those are mostly "read" errors, as a single 'write" error will take a drive out-of-service.

however i'm really worried that i might have problems with the data on my array.

You do have issues...

I have tryed to re-sync the parity a few times now and it seems to go well each time until it gets to the end where it fails. What % im not sure because i haven't seen the last 10% of the parity build, but before that there are no errors and everything seems fine.

OK, that indicates the errors are in the higher numbered sectors on the drive.

I have tryed changing sata ports and changing sata cables on the parity drive and the samsung drive.

That would not affect failing hard disks.

I have run a short smart test on the 2 drives. they both passed, but the parity drive had a raw value of 99 for the Current_Pending_Sector. I read this is bad but what do it mean I should do?

You are right, it is bad.  It indicates the disk has 99 sectors where the checksum at the end of the sector does not match the data on the  sector.  It cannot be fixed other than by re-writing the sector.

 

I really have no idea what to do, so any help would be really appreciated. I'm really don't like having a unprotected array its kinda scary!

Since your parity disk keeps failing, I'd start by ordering a replacement.

My syslog got reset this morning when i turned off the array and restarted it so i'm doing another parity sync to get a sys log after the sync fails and after i get those error numbers i mentioned, and ill post that as soon as its done. However I have attached the short smart test results for parity drive and the one in the array that gets numbers in the errors column after an attempt to resync parity.

any advice i would greatly appreciate

 

edit: im trying to run the parity syn again but it running at 300~ kb/sec normally its ~115 mb /sec. any idea what this could be??

yes, the disk is failing, and constantly slowing things down.

 

Only way to know is to look in the syslog.  The older syslog is probably still there (unless you rebooted) as /var/log/syslog.1

You really need to attach to your next post a syslog.

 

Joe L.

this error on the parity drive usually indicates either a poor quality SATA cable, or one that is picking up noise from an adjacent cable, or a noisy power supply voltage.

199 UDMA_CRC_Error_Count    0x0032  200  200  000    Old_age  Always      -      1

 

If you've neatly tie-wrapped all the sata cables in a bundle, you've made them most likely to pick up induced noise.    remove the tie-wraps.  Let the cables have some distance between them.  Unless you have shielded SATA cables, you've created the issue by letting the cables induce noise in each other.  (distance between them is your friend)

first off thank you for replying so quickly and with such a detailed a response for what little info i could give. you obviously know your stuff and i appreciate the help.

 

Those are mostly "read" errors, as a single 'write" error will take a drive out-of-service.

 

good to know, i would hate to think that samsung drive is failing too

 

You are right, it is bad.  It indicates the disk has 99 sectors where the checksum at the end of the sector does not match the data on the  sector.  It cannot be fixed other than by re-writing the sector.

 

I am getting a new WD 3TB Red drive tomorrow to replace it, but can this drive be reused at all? it is still under warranty, will i be able to get a refund because of this? you mention rewriting the drive would that be like doing another preclear or is it like a low level format then preclear?

 

Only way to know is to look in the syslog.  The older syslog is probably still there (unless you rebooted) as /var/log/syslog.1

You really need to attach to your next post a syslog.

 

unfortunetly i did reboot when trying to trouble shoot by changing/repluging the sata cables, but i've attached a syslog of the slow attempt at parity sync, stoped the sync after 5 mins (about 400kb/s speed) hopefully that can tell you something

 

this error on the parity drive usually indicates either a poor quality SATA cable, or one that is picking up noise from an adjacent cable, or a noisy power supply voltage.

199 UDMA_CRC_Error_Count    0x0032  200  200  000    Old_age  Always      -      1

 

If you've neatly tie-wrapped all the sata cables in a bundle, you've made them most likely to pick up induced noise.    remove the tie-wraps.  Let the cables have some distance between them.  Unless you have shielded SATA cables, you've created the issue by letting the cables induce noise in each other.  (distance between them is your friend)

 

I did actually try and neaten the cables up, and i did use a molex to sata power adapter i got off ebay for the parity. if this could have caused the problem is the drive issue fixable? I will be much more careful about cabling from now on!

syslog.txt

I am getting a new WD 3TB Red drive tomorrow to replace it, but can this drive be reused at all? it is still under warranty, will i be able to get a refund because of this? you mention rewriting the drive would that be like doing another preclear or is it like a low level format then preclear?

You can try once more to put this existing disk back into service, but without the older syslog, we don't know what caused the "write" to the disk to fail the past two times.

 

Odds are it will just fail again unless you corrected something.  (loose connection, bad cable, etc)

 

With as many un-readable sectors I'd just RMA the disk.    If you already have a disk ordered as a replacement, then when you get the new disk, run it through a preclear cycle (just to stress test it) and then assign it as your parity drive.

I am getting a new WD 3TB Red drive tomorrow to replace it, but can this drive be reused at all? it is still under warranty, will i be able to get a refund because of this? you mention rewriting the drive would that be like doing another preclear or is it like a low level format then preclear?

You can try once more to put this existing disk back into service, but without the older syslog, we don't know what caused the "write" to the disk to fail the past two times.

 

Odds are it will just fail again unless you corrected something.  (loose connection, bad cable, etc)

 

With as many un-readable sectors I'd just RMA the disk.    If you already have a disk ordered as a replacement, then when you get the new disk, run it through a preclear cycle (just to stress test it) and then assign it as your parity drive.

 

Just a quick follow up post. You were right, the Disk was failing. I took it to the computer store I got it from and they tech guy did a quick test and said it was pretty much completely dead. They gave me a full refund, which i then used to buy a Red WD 3TB drive for the same price!

 

Installed the red drive as parity after doing 2 pre clears, everything is working perfectly now.

 

Thanks alot Joe