SOLVED server down with one missing disk and one dead disk

[jpimlott]

My unraid server lost disk4 in the middle of the night. When I stopped and restarted the server a second disk failed.

Disk 8 the second disk failure does not past POST. I have put a new precleared disk in the system to replace disk8.

Disk 4 passes post and passes smart test. I would like to know if there is a way to reset disk4 to make it valid again.

I would then be able to rebuild disk8 from parity.

If that is not possible  my next step would be to copy off data off disk 4 and save it to another disk.

that is not the est option as I would then loose all of disk8 because I would loose all of the data on disk8

[Joe L.]

Yes, you can set disk4 as invalid and hopefully then recover disk8.

 

The secific commands differ for unraid versions. You dd not supply details or a syslog, so nobody can give you specifics

[jpimlott]

its is 4.7  I tried to post the syslog but it not show

here it is

syslog-2011-12-29.txt

[Joe L.]

its is 4.7  I tried to post the syslog but it not show

here it is

 

On version 4.7, you should be able to

 

log in via telnet or on the system console

type

initconfig

then, answer "Yes" to its prompt. (capital "Y", lower case "es")

 

then type:

/root/mdcmd set invalidslot 8

 

Next, refresh the management web-page.  It should show disk8 as red, the others as green.

Then, start the array.  It should reconstruct disk8.    (you should be able to press "Refresh" to follow the progress)

 

If you do not see "writes" to disk 8, press "Stop" as quickly as possible.  (It should be reading from all disks other than 8, and writing to 8.)

 

NOTE: these instructions will NOT work as described for any of the unRAID 5.X series.

 

Joe L.

[jpimlott]

Thanks for the info i will try when I get home.

 

[jpimlott]

its is 4.7  I tried to post the syslog but it not show

here it is

 

On version 4.7, you should be able to

 

log in via telnet or on the system console

type

/root/mdcmd initconfig

then, answer "Yes" to its prompt. (capital "Y", lower case "es")

 

then type:

/root/mdcmd set invalidslot 8

 

Next, refresh the management web-page.  It should show disk8 as red, the others as green.

Then, start the array.  It should reconstruct disk8.    (you should be able to press "Refresh" to follow the progress)

 

If you do not see "writes" to disk 8, press "Stop" as quickly as possible.  (It should be reading from all disks other than 8, and writing to 8.)

 

NOTE: these instructions will NOT work as described for any of the unRAID 5.X series.

 

Joe L.

Thank you for the help. It Worked and and the server is rebuild disk8 now. I did not need the

mccmd part of the command .

It made a few write to parity and disk4  when it first started then went to only write to disk 8.

 

I owe you a beer or 2 or 3 ...

john

 

 

[Joe L.]

You are correct.  I was wrong.  It should have only been

initconfig

not

/root/mdcmd initconfig

 

Joe L.