Jump to content

5b12a running awesome for 2 months suddenly starts throwing parity errors


Recommended Posts

(syslog attached)


This unraid box is nothing but insurance.  It does a backup of my main server which is also running b12a without any issue.  Yesterday it was doing an rsync backup of the main server when I noticed parity errors after the server sync.  Puzzled, I just shut it down and restarted and had it start a parity check without correcting parity.  It immediately found 2 errors, but was only running at 2mb/sec rather than the usual 100mb/sec so I stopped it.  I started it up again and now it immediately found 14 errors while checking parity (without correcting) and it is running at the slow speed.  Finally it stops with disk2 red balled after finding 1152 errors.


Nothing has changed on this server in several months but I know the ps was marginal when I built this server out of junk laying around.  (no worries here as it is just a rotating backup server that gets switched out for another every few weeks)


Asus m4a 785 m

AMD 140

2 gig RAM

5 - 2 tb drives

Seasonic 330W ps with split 12v rails (yeah I know I shouldn't have....)


I'm thinking this is not a b12a issue but rather something simple like bad or marginal hardware.  Do you agree?


Link to comment

It has some Samsung F4 drives but they all have the firmware updated.  Here is the drive summary, disk2 presently red balled and it's a Seagate.

Device 	Identification 	Temp. 	Size 	Free 	Reads 	Writes 	Errors
parity 	SAMSUNG_HD204UI_S2H7J1BZA27926 (sdb) 1953514552 	17°C 	2 TB 	- 	145 	45 	0
disk1   ST32000542AS_5XW18FW3 (sda) 1953514552 	20°C 	2 TB 	252.85 GB 	164 	13 	0
disk2   ST32000542AS_5XW1CSPD (sdc) 1953514552 	18°C 	2 TB 	361.16 GB 	53 	0 	0
disk3   SAMSUNG_HD204UI_S2H7J1BZA27924 (sdd) 1953514552 	17°C 	2 TB 	312.41 GB 	154 	10 	0
disk4  	SAMSUNG_HD203WI_S1UYJ1YZ601695 (sde) 1953514552 	17°C 	2 TB 	1.62 TB 	153 	10 	0
flash   DataTraveler_2.0 	- 	1.03 GB 	792.15 MB 	157 	8 	0



What are the makes and models of the hard drives.  Samsung's F4 drives initially shipped with buggy firmware that would cause parity problems.  (It wrote blocks of 0 to the drive instead of the actual data.)

Link to comment

Replaced the PS with a Corsair 430 and the system happily rebuilt the red-balled drive and is now in the middle of a parity check.  If this completes successfully, I guess I only have to figure out if the drive might have corrupt data on it from the failure mode. 


How do you check for corrupted data after a rebuild?

Link to comment


This topic is now archived and is closed to further replies.

  • Create New...