Hi All

 

I have just received a dreaded red ball on disk 8 of array. I have attached the syslog and smartctl report. Is my drive dead or dying?

 

smart.txt

syslog.txt

I have run an extended smart disk check and found no errors. Is it safe to activate the drive again?

I have run an extended smart disk check and found no errors. Is it safe to activate the drive again?

it is PROBABLY safe to activate the drive again. 

 

having said that, if by any chance you have a spare drive the recommendation would be to rebuild onto that spare drive putting the drive currently marked as failed somewhere safe for the time being.  That would minimise any chance of data loss as you still have the ostensibly 'failed' disk in its current state as a fallback.    If you go down that route the recommendation once you are happy data recovery has completed successfully the 'failed' drive can then be put through a pre-clear cycle to provide a confidence check.

So I rebuilt the drive and everything was fine for a few days and then errors again. Surely when rebuilding the drive it should find problems then? Or is it because unraid is now trying to write to free space at the end of the drive and encountering the bad sectors there? Although, I am sure the unused sectors have 0's written to them during rebuild?

Here is the syslog from when the problem started. Problem with the Card?

 

 

 

 

 

Jun 30 19:53:25 Tower kernel: md: disk8 write error, sector=934694128

Jun 30 19:53:25 Tower kernel: md: disk8 write error, sector=934694136

Jun 30 19:53:25 Tower kernel: md: disk8 write error, sector=934694144

Jun 30 19:53:25 Tower kernel: md: disk8 write error, sector=934694152

Jun 30 19:53:25 Tower kernel: md: disk8 write error, sector=934694160

Jun 30 19:53:25 Tower kernel: md: disk8 write error, sector=934694168

Jun 30 19:53:25 Tower kernel: md: disk8 write error, sector=934694176

Jun 30 19:53:25 Tower kernel: md: disk8 write error, sector=934694184

Jun 30 19:53:25 Tower kernel: md: disk8 write error, sector=934694192

Jun 30 19:53:25 Tower kernel: md: disk8 write error, sector=934694200

Jun 30 19:53:25 Tower kernel: md: disk8 write error, sector=934694208

Jun 30 19:53:25 Tower kernel: md: disk8 write error, sector=934694216

Jun 30 19:53:25 Tower kernel: md: disk8 write error, sector=934697984

Jun 30 19:53:30 Tower kernel: sas: sas_form_port: phy0 belongs to port0 already(1)!

Jul 1 00:00:01 Tower kernel: mdcmd (50): check

Jul 1 00:00:01 Tower kernel: md: recovery thread woken up ...

Jul 1 00:00:01 Tower kernel: md: recovery thread has nothing to resync

Jul 1 06:45:30 Tower dhcpcd[1098]: eth0: renewing lease of 192.168.2.52

Jul 1 06:45:30 Tower dhcpcd[1098]: eth0: acknowledged 192.168.2.52 from 192.168.2.1

Jul 1 06:45:30 Tower dhcpcd[1098]: eth0: leased 192.168.2.52 for 259200 seconds

Jul 1 06:45:30 Tower ntpd[22068]: ntpd exiting on signal 1

Jul 1 06:45:31 Tower ntpd[25735]: ntpd [email protected] Sat Apr 24 19:01:14 UTC 2010 (1)

Jul 1 06:45:31 Tower ntpd[25736]: proto: precision = 0.567 usec

Jul 1 06:45:31 Tower ntpd[25736]: ntp_io: estimated max descriptors: 40960, initial socket boundary: 16

Jul 1 06:45:31 Tower ntpd[25736]: Listen and drop on 0 v4wildcard 0.0.0.0 UDP 123

Jul 1 06:45:31 Tower ntpd[25736]: Listen normally on 1 lo 127.0.0.1 UDP 123

Jul 1 06:45:31 Tower ntpd[25736]: Listen normally on 2 eth0 192.168.2.52 UDP 123

Jul 1 12:25:25 Tower sSMTP[5683]: Creating SSL connection to host

Jul 1 12:25:26 Tower sSMTP[5683]: SSL connection using RC4-SHA

Jul 1 12:25:30 Tower sSMTP[5683]: Sent mail for [email protected] (221 2.0.0 closing connection eo4sm41781372wid.4 - gsmtp) uid=0 username=root outbytes=422

Jul 1 12:30:23 Tower kernel: program smartctl is using a deprecated SCSI ioctl, please convert it to SG_IO

Jul 1 12:30:23 Tower last message repeated 6 times

Jul 1 12:30:24 Tower sSMTP[9559]: Creating SSL connection to host

Jul 1 12:30:24 Tower sSMTP[9559]: SSL connection using RC4-SHA

Jul 1 12:30:29 Tower sSMTP[9559]: Sent mail for [email protected] (221 2.0.0 closing connection wi9sm47295497wjc.23 - gsmtp) uid=0 username=root outbytes=8911

Jul 1 12:30:51 Tower kernel: program smartctl is using a deprecated SCSI ioctl, please convert it to SG_IO

Jul 1 12:30:51 Tower last message repeated 6 times

Jul 1 12:30:52 Tower sSMTP[10197]: Creating SSL connection to host

Jul 1 12:30:52 Tower sSMTP[10197]: SSL connection using RC4-SHA

Jul 1 12:30:57 Tower sSMTP[10197]: Sent mail for [email protected] (221 2.0.0 closing connection gq4sm7977568wib.8 - gsmtp) uid=0 username=root outbytes=8936

It is difficult to guess the cause without seeing what lead up to that error.

 

What does seem clear is that it has dropped offline for some reason.    This can have a wide variety of causes:

- cabling issues (power of SATA)

- disk controller problem

- power supply problem

- disk drive issue

 

Thanks for the reply. I will start tonight by reseating the cabling and see if that helps. Hope its not the controller! Will follow up with my findings.

This has just popped up on the syslog

 

Jul 1 12:32:00 Tower kernel: program smartctl is using a deprecated SCSI ioctl, please convert it to SG_IO

Jul 1 12:33:04 Tower last message repeated 7 times

Jul 1 12:34:07 Tower last message repeated 7 times

Jul 1 12:35:11 Tower last message repeated 7 times

Jul 1 12:36:14 Tower last message repeated 7 times

Jul 1 12:37:18 Tower last message repeated 7 times

Jul 1 12:38:21 Tower last message repeated 7 times

Jul 1 12:39:25 Tower last message repeated 7 times

Jul 1 12:40:28 Tower last message repeated 7 times

Jul 1 12:41:32 Tower last message repeated 7 times

Jul 1 12:42:36 Tower last message repeated 7 times

Jul 1 12:43:39 Tower last message repeated 7 times

Jul 1 12:44:43 Tower last message repeated 7 times

Jul 1 12:45:46 Tower last message repeated 7 times

Jul 1 12:46:50 Tower last message repeated 7 times

Jul 1 12:47:53 Tower last message repeated 7 times

 

Problem with card maybe?