My current unRAID server works perfectly well as a file server, but with a fairly low powered AMD CPU and 8GB of RAM its not good for more than a few Docker containers.  I would like to move my drives, controller and USB key to a Dell PowereEdge c2100 that I have in my rack. It has 2 6-core Xeons and 72 GB of DDR3 RAM so it should be able to handle everything I throw at it.  A few months ago I tried to move my unRAID installation to this server as a guest in ESXi.  The results were a disaster and have been chronicled in this thread:

 

http://lime-technology.com/forum/index.php?topic=40458.0

 

TL;DR. I moved the drives and USB key back to my original server, ran reiserfsck on one of the disks and have had no problems since. I have since stopped using the Dell server for ESXi and would like to run unRAID natively. I have a spare USB key, so I have put spare drives in all 12 slots, booted unRAID, preacleared all drives with 3 cycles (with 0 faults) and created a test array. I then filled the array about half full, powered it down, shuffled the drives around then rebooted into unRAID and restarted the array.  Everything mounted fine, so I copied the data that I had written off the array just to see how it would perform and it worked fine.

 

Given my past history with this server though, I am a bit nervous about putting the drives from my main unRAID server into this chassis.  Is there anything else I can do to test and ensure that I won't have any problems when I move everything over?

So is nobody chiming in because I am being too nervous? I don't want to lose any data and am nervous given my past experience trying to move drives to different hardware.

If you already have drives in the new server why not  copy the data over the network from the existing server to the new server?

It's just a bunch of small drives for testing. Mostly 160gb. The drives in my production server are all 4tb, plus my cache pool and docker containers.

I recently moved my array to a new MB and CPU and had no issues. Powerdown. Replaced HW. Powered up.

 

You can configure the array to not autostart so when you power up the first time you can make sure that all assignments are correct and only then hit the start button.

That's the procedure I followed last time and plan on following this time. If I start the array in maintenance mode are there tests besides smart reports I can run on the disks?

You are going to remove everything from your test, including the USB key, and just go with the drives from your production server with its USB key? If so it should be fine. Before you shut down the production server just set it to not autostart like theone suggested. I don't see any reason to do any drive tests after the move. Might as well do them before the move if you think it makes sense for some reason. I have changed my hardware twice and it always started fine.

Yes, that's the idea. I will give this a shot tomorrow. I will also move my IBM1015 controller since it had been working well for me so far. Fingers crossed that all goes well.

 

You are going to remove everything from your test, including the USB key, and just go with the drives from your production server with its USB key? If so it should be fine. Before you shut down the production server just set it to not autostart like theone suggested. I don't see any reason to do any drive tests after the move. Might as well do them before the move if you think it makes sense for some reason. I have changed my hardware twice and it always started fine.

 

AND take a screenshot of the assignments it helps

 

 

AND take a screenshot of the assignments it helps

 

Will do.

 

Just moved the drives + USB key to the other server and powered it up. Upon logging into the Web GUI the SSDs were not assigned to the cache slots, so I assigned them to the same slots as the screenshot I took prior to powering down.  All other drive assignments were correct.  I started the array and the shares look correct, including my cache share. But Docker will not start.

 

I see the following in the logs.

 

Nov 20 09:52:47 NalboneFS1 logger: Warning: stream_socket_client(): unable to connect to unix:///var/run/docker.sock (No such file or directory) in /usr/local/emhttp/plugins/dynamix.docker.manager/include/DockerClient.php on line 490

 

Any thoughts? Syslog attached.

nalbonefs1-syslog-20151120-0949.zip

Just moved the drives + USB key to the other server and powered it up. Upon logging into the Web GUI the SSDs were not assigned to the cache slots, so I assigned them to the same slots as the screenshot I took prior to powering down.  All other drive assignments were correct.  I started the array and the shares look correct, including my cache share. But Docker will not start.

 

I see the following in the logs.

 

Nov 20 09:52:47 NalboneFS1 logger: Warning: stream_socket_client(): unable to connect to unix:///var/run/docker.sock (No such file or directory) in /usr/local/emhttp/plugins/dynamix.docker.manager/include/DockerClient.php on line 490

 

Any thoughts? Syslog attached.

 

Apparently this has happened to a few people and removing the Docker image and creating a new one tends to resolve the issue. I have done so and so far have recreated my MythTV docker which appears to be working fine.  Will try with the others and report back.

Something is wrong with my cache pool. The GUI reports that it is healthy, but the system logs show a bunch of read-only filesystem errors. What do I do? Syslog attached.

nalbonefs1-syslog-20151121-0655.zip