Is it safe?

[jag131990]

Hi guys,

I apologize if this topic has been covered - I am interested to know is it safe to continue to use FreeNAS (read heavily utilize) in the event the a disk is failing (read not completely failed yet).

Background: I have a disk reporting SMART errors and error count is increasingly daily.. I believe this disk is imminently failing given the emails I get daily from FreeNAS.

Code:

SMART Attributes Data Structure revision number: 16
Vendor Specific SMART Attributes with Thresholds:
ID# ATTRIBUTE_NAME		  FLAG	 VALUE WORST THRESH TYPE	  UPDATED  WHEN_FAILED RAW_VALUE
  1 Raw_Read_Error_Rate	 0x002f   200   199   051	Pre-fail  Always	   -	   19212
  3 Spin_Up_Time			0x0027   187   179   021	Pre-fail  Always	   -	   5641
  4 Start_Stop_Count		0x0032   085   085   000	Old_age   Always	   -	   15388
  5 Reallocated_Sector_Ct   0x0033   193   193   140	Pre-fail  Always	   -	   222
  7 Seek_Error_Rate		 0x002e   200   200   000	Old_age   Always	   -	   0
  9 Power_On_Hours		  0x0032   072   072   000	Old_age   Always	   -	   20789
10 Spin_Retry_Count		0x0032   100   100   000	Old_age   Always	   -	   0
11 Calibration_Retry_Count 0x0032   100   100   000	Old_age   Always	   -	   0
12 Power_Cycle_Count	   0x0032   100   100   000	Old_age   Always	   -	   979
192 Power-Off_Retract_Count 0x0032   200   200   000	Old_age   Always	   -	   38
193 Load_Cycle_Count		0x0032   195   195   000	Old_age   Always	   -	   15349
194 Temperature_Celsius	 0x0022   108   095   000	Old_age   Always	   -	   42
196 Reallocated_Event_Count 0x0032   023   023   000	Old_age   Always	   -	   177
197 Current_Pending_Sector  0x0032   200   199   000	Old_age   Always	   -	   3
198 Offline_Uncorrectable   0x0030   100   253   000	Old_age   Offline	  -	   0
199 UDMA_CRC_Error_Count	0x0032   200   200   000	Old_age   Always	   -	   0
200 Multi_Zone_Error_Rate   0x0008   200   197   000	Old_age   Offline	  -	   65

My real question is regards to data parity calculations and the meta data that is striped across my array (RAIDZ2) given this disk is unreliable...

I have been pandering to my FreeNAS' email alerts and have paused all downloading and limit my use to basic 'reads only' of my data on this system while I await a replacement disk. I feel (perhaps unjustified) that it would be safer to use this system with the disk completely removed and the system downgraded such that this disk is not going to further cause havoc in the case of the parity and metadata associated with creating new files on this array.

Is there any literature you could share with me on this, or experience you may have to share?

 

[Pitfrr]

I have no litterature to propose but I'd say it's not a problem to use your system with a degraded disk, that's the whole purpose of RAIDZx!

You said you ordered a replacement drive, that's fine and in the meantime you can use your system normally (even though I'd monitor closely the other drives to prevent a second drive failure...).
Once the new drive is arrived just replace it (see in the documentation to follow the procedure) and you're done.

Of course, I hope you have a backup of your data (otherwise it would be advised to back it up right away! ;-O).

 

[jag131990]

Thanks for your response, I understand that its the purpose of RAIDZx, however I guess my reservations stem from the fact that this disk is now in unknown territory and I don't understand what metadata stripe creation looks like in the event a disk can be partially working and co-existing in this process vs just fully removed.

I have been trying to sort out a backup solution today however have had to disappointingly revert to a slow network copy to a portable HDD when I learnt that Crashplan no longer works! Hopefully all is well. Thanks again.

 

[Ericloewe]

Please use [CODE][/CODE] tags whenever you paste console output. SMART data is a pain to read when the formatting is messed up.

 

[wblock]

is it safe to continue to use FreeNAS (read heavily utilize) in the event the a disk is failing (read not completely failed yet).

It's not ideal, some redundancy will go away when that drive actually fails. But it will work. Just replace the failing drive as quickly as possible. Maybe get a spare to keep on-hand or even as a hot spare.

The reallocated sectors happen when the drive writes a block, reads it back, and finds the data does not match. It marks the bad sector, allocates an existing spare sector, and uses that instead. So far, the drive is coping, but 222 reallocations is a lot.

 

[jag131990]

Yeah does seem like a fair bit..

Also get emails like this last one:

Code:

Device: /dev/da7 [SAT], Self-Test Log error count increased from 2 to 3
Device: /dev/da7 [SAT], 3 Currently unreadable (pending) sectors

 

[Deleted47050]

Yeah that disk is failing, but you can continue using the system. As other people said already, just replace it as soon as you get the new drive and you will be good.

In a freenas test system I use, I have a disk with a couple of unreadable pending sectors that has been that way for more than a year. Since it's a test system, I don't care about the data or the drive, but it's just to show that you can continue using freenas anyway until you get your replacement.


Sent from my iPhone using Tapatalk