I have a drive that is possibly going bad. I have an 11 drive system with only about %50 being used. Instead of replacing the drive right away I would like to just remove it and let unraid work that way until I eventually buy a new drive. Is there a procedure to do this? I know if I remove the drive the unraid system will still operate. At this point where would I go to let unraid know I'm not going to reinstall the removed drive? It is probably an easy procedure but I won't rely on my unraid knowledge at all compared to the experts here.

 

Thank you- Mike.

 

basically,

 

you copy the data off the drive you want to remove,

stop the array,

pull the drive you want to discard,

then initconfig.

you then start/rebuild the array without that one drive.

 

all the usual warnings about screen capturing your array, make sure you know what drive is parity, do not format any drives if it asks you to

it would be best if you do a parity check first.

 

see http://lime-technology.com/wiki/index.php?title=FAQ#How_do_I_remove_a_hard_disk_that_I_do_not_plan_on_replacing.3F

 

What if I don't or can't copy anything off the drive I want to remove? I don't think the copy will make it properly, so I just want to remove the drive soon as possible and then let unraid "rebuild" itself without the missing drive. Right now I removed the drive. Booted up the server and it says disk3 is missing, which is correct. I have not done anything else. I'm tempted to goto devices and UNassign it, but I'm not sure. There are no "directions" for this and the wiki and FAQ's are hard to interpret.

 

 

What if I don't or can't copy anything off the drive I want to remove? I don't think the copy will make it properly, so I just want to remove the drive soon as possible and then let unraid "rebuild" itself without the missing drive. Right now I removed the drive. Booted up the server and it says disk3 is missing, which is correct. I have not done anything else. I'm tempted to goto devices and UNassign it, but I'm not sure. There are no "directions" for this and the wiki and FAQ's are hard to interpret.

 

 

Stop the array

log in via telnet, type

initconfig

respond with "Yes" (capital "Y", lower case "es")

go back to the main web-management page and start the array.  Parity will be newly calculated on the new array configuration.

Thanks. This is what I had in my head to do but didn't trust myself. Since it couldn't hurt I reinstalled the disk3 and trying to copy what I can over to the other drives. Since I just use ONE usershare to access all my stuff I hope this doesn't mess anything up when I'm all done. I guess that's what unraid does, lets us view our shares that look as one folder across multiple drives.

 

If I can't finish the copy from my disk3 to the other drives, how would I finish this up? Some files were moved over and some are still on disk3. I'm getting too many read errors and the speed dropped down to like 10K/sec. Still the same procedure?

NO,

 

you will loose the data on disk 3.

I dont know what the procedure is to remove a blank drive off the back of your array and then have it rebuild disk3 on that  disk you removed....

I know there is a way to remove a disk that you know is blank and tell it parrity is still good. I am not sure if that is the route to go..

 

I think joe  missed you wanted to keep the data.

 

it was not clear in the first post that you had data on the failing disk. we both assumed it was a thow-away

I thought the logical thought would be there is data on a disk that is failing. Why would I keep an empty disk in the array? Anyway, even with the read errors I'm still copying files. I'm down to the last GB's and the files I'm moving now are just movie files that are important but not critical. So, I'll end up with a blank disk3 which then I'll remove from the array and then do an initconfig from the CL. I'll eventually buy another couple drives but I have 5TB available now which should hold me over for a while. All the hard drives I usually order are too expensive now or out of stock. Since unraid stable doesn't support 3TB drives I ponder the idea of buying a Synology 1511 and adding 2 extra bays. At least i'll have super speeds reading and writing, plus 3TB support on a fully protected system.

 

 

Any hoot I was able to move everything off the "bad" drive. Speeds were horrible since there were so many read errors but everything moved without error. Unassigned the drive, rebooted and parity started a rebuild. Rebuild is over, went with no problems and I'm back to normal minus one drive. Still 5TB free. Right now I'm trying to figure out IF I had the Synology 1511+ and had the same thing happen what would have been the resolution or procedure? Could it even be done? Maybe someone that knows those systems can chime in?

 

Thanks- Mike

 

keep in mind, there was a simple fix...replace the drive, let it rebuild.

 

whenever you run any sort of drive array, you should always have a spare on hand.

this was just complicated because you did not want to replace the drive.

 

I understand your reasoning, don't get me wrong. I am not trying to single you out. One thing i do see over and over in unraid. people over commit their drives.

 

with unRAID, I see people buy 15 drives at once and slap all 15 drives into the array at once.

if you only need 6 drives at first.. I would recommend only plugging in 8 drives, preclear the rest and leave them out of the array. unplugged if possible. then add them as you need them.

also, if you have a 15 drive array, buy 16 drives and leave the spare with the system somewhere ready to go. I really recommend keeping a spare (equal to your parity in size) on hand at all times. use that spare for cache if you want (altho it looses its preclear report).

 

for example, i have 16x 3TB drives for my primary array, but i only have 8 online. they rest are precleared and on standby.

this would help prevent problems like yours.

 

with a real raid, this is not always the case since it can take days to expand an array. we usually wait for holidays to expand arrays with heavy policing in the meantime.

bad drives happen. being ready for them.

 

In your case, you could have just swapped drives and not thought about it. Perhaps this might help the next person.

 

I'm glad you got it worked out in the end. i hate loosing data..