read errors - how screwed am I

[daan_SVK]

hi there, 

 

Fix Common Issues is telling me my Disk1 has read errors, I ran the extended SMART diagnostics and it came back clear, however the disk does show 200 relocated sectors. 

 

Server diagnostics is attached, how bad is it?

 

I have already ordered a replacement drive, just want to make sure the other drives are not having issues and I can rebuild from parity. 

 

thanks a lot

tower-diagnostics-20180504-1302.zip

[JorgeB]

You should have no issues replacing it, and I would give it a second chance, since SMART looks fine and the extended test completed without error.

[daan_SVK]

ok, thanks. I have already ordered a new WD RED to replace the old Green. 

 

what do you mean by giving it a second chance, you mean to put it back into the array after successfull pre-clearing?

 

when it comes to replacing the failing drive, do i need to pre-clear the new disc? 

 

thanks again for your feedback

[JorgeB]

15 hours ago, daan_SVK said:

what do you mean by giving it a second chance, you mean to put it back into the array after successfull pre-clearing?

No, just keep using it, sometimes those type of errors happen only once, but it if does happen again then replace the disk.

[JorgeB]

15 hours ago, daan_SVK said:

when it comes to replacing the failing drive, do i need to pre-clear the new disc? 

Don't need to but it's good to test the new disk before using it, with preclear or any other disk test util.

[daan_SVK]

Don't need to but it's good to test the new disk before using it, with preclear or any other disk test util.

Ok, thanks.

I think I'll replace the flagged WD Green with the RED I ordered. Once rebuild, I'll preclear the Green and if successful and SMRT comes back clean, I'll pop it back into the array.

Sent from my ONEPLUS A3000 using Tapatalk

[Frank1940]

23 hours ago, daan_SVK said:

I'll preclear the Green and if successful and SMRT comes back clean, I'll pop it back into the array.

 

 

I hope you are not proposing to swap out that 'Red' that you just put in.  If you need more space now, I could see adding it as a 'new' data disk.  Otherwise, keep it around as a spare until you have reason to replace another disk.  (Rebuilding a disk always has a potential for something going wrong. The best advice has always been-- don't screw with a working system just because you can.) 

[daan_SVK]

 
 
I hope you are not proposing to swap out that 'Red' that you just put in.  If you need more space now, I could see adding it as a 'new' data disk.  Otherwise, keep it around as a spare until you have reason to replace another disk.  (Rebuilding a disk always has a potential for something going wrong. The best advice has always been-- don't screw with a working system just because you can.) 

I didn't put in anything yet. I have a new RED in the mail that will be replacing the flagged GREEN.

Based on the previous comment I was considering preclearing the GREEN once the array has been rebuild successfully with the new RED.

My understanding after reading the comments was that since the SMART report was clean, the GREEN might be reused if preclear comes back as successful.

Please correct me if I'm under wrong assumptions here

Sent from my ONEPLUS A3000 using Tapatalk

[Frank1940]

We seem to have a problem in communication.  My comment about the Green Drive was not to say not to use it.  That is your choice.  Putting the new Red Drive into the array and that should restore the array to normal operation.   You can then preclear the Green drive and if you find that there are no errors in the preclear and the SMART report, you can, at that point, decide if you want to continue to use it.   

 

The point I was trying to make was that you should not simply remove a working drive from the array to simply replace it with that Green Drive.  Wait until you have another issue.  If you decide to add another disk to the array, you are free to use that Green Drive.  NEVER do a disk rebuilt unless you have a valid reason for doing it. When I read this comment:

On 5/6/2018 at 11:14 AM, daan_SVK said:

I'll preclear the Green and if successful and SMRT comes back clean, I'll pop it back into the array.

 

I assumed that you were going to immediately (and with any thought or reason) use it to replace one of the other current drives in the array.  

 

PS-- I downloaded your diagnostics file and looked at disk1.  (I found it hard to believe that @johnnie.black would say that a disk with 200 reallocated sectors "looks fine".) The SMART report on that disk look clean to my eyes.  You do not have Reallocated_Sectors_Ct of 200.  You have zero!  (You should be looking at the RAW_VALUE for the attributes (5, 187, 188, 197, 198, and 199) that most folks check when evaluating disk health.) From the attributes table, you have an absolutely good disk.

 

 

[daan_SVK]

12 minutes ago, Frank1940 said:

We seem to have a problem in communication.  My comment about the Green Drive was not to say not to use it.  That is your choice.  Putting the new Red Drive into the array and that should restore the array to normal operation.   You can then preclear the Green drive and if you find that there are no errors in the preclear and the SMART report, you can, at that point, decide if you want to continue to use it.   

 

The point I was trying to make was that you should not simply remove a working drive from the array to simply replace it with that Green Drive.  Wait until you have another issue.  If you decide to add another disk to the array, you are free to use that Green Drive.  NEVER do a disk rebuilt unless you have a valid reason for doing it. When I read this comment:

 

I assumed that you were going to immediately (and with any thought or reason) use it to replace one of the other current drives in the array.  

 

PS-- I downloaded your diagnostics file and looked at disk1.  (I found it hard to believe that @johnnie.black would say that a disk with 200 reallocated sectors "looks fine".) The SMART report on that disk look clean to my eyes.  You do not have Reallocated_Sectors_Ct of 200.  You have zero!  (You should be looking at the RAW_VALUE for the attributes (5, 187, 188, 197, 198, and 199) that most folks check when evaluating disk health.) From the attributes table, you have an absolutely good disk.

 

 

 

sorry for the misunderstanding and thanks for taking the time to review the logs, I really appreciate it. 

 

my daily health checks are failing due to what's happening with this disc. Should I be just ignoring the error?

 

[Frank1940]

2 minutes ago, daan_SVK said:

my daily health checks are failing due to what's happening with this disc. Should I be just ignoring the error?

 

 

No, do not ignore the disk. Replace that disk with the new one, allowing it to rebuilt, should fix the problem. 

 

Once you do that unRAID will 'forget' the problems that has had with that disk.  But run the preclear checks on it.  (I, personally, would run three full preclear cycles but then I am risk adverse and would rather be safe than sorry.) 

 

[daan_SVK]

1 hour ago, Frank1940 said:

 

No, do not ignore the disk. Replace that disk with the new one, allowing it to rebuilt, should fix the problem. 

 

Once you do that unRAID will 'forget' the problems that has had with that disk.  But run the preclear checks on it.  (I, personally, would run three full preclear cycles but then I am risk adverse and would rather be safe than sorry.) 

 

 

thanks again for your input. 

 

It appears I got shipped a 4TB instead of 3TB and my parity drive is 3TB only. 

 

any idea if this is still valid?

https://lime-technology.com/wiki/The_parity_swap_procedure

 

that would allow me to upgrade to a new parity drive and use the old one to replace my failing data drive. 

Edited May 7, 2018 by daan_SVK

[Frank1940]

Yes.  (I would print out a copy of the instructions and read them until you completely understand them.  Figure exactly what the state of your array and  which steps are required and highlight them.  Then as you do each step, check it off.)  

[JorgeB]

9 hours ago, daan_SVK said:

my daily health checks are failing due to what's happening with this disc. Should I be just ignoring the error?

Rebooting the server will clear those errors.