Reallocate disks

It's tricky, but should be possible assuming that all of your filesystems can fit entirely on the new 5 NL-SAS disks (not including the existing NL-SAS disks already allocated).  If your filesystems will not fit entirely on the new NL-SAS disks, this method will not work.

You would build a separate storage pool entirely out of the new disks and allocate them to the data movers.  If AVM makes a new pool for them (which it may since they are all NL-SAS disks), then it's a bit easier.  If it doesn't, then you will have to create your own file pool using those new LUNs so that they are separate. 

Now that you have 2 file pools, you will create new empty filesystems on them of the same size as the originals.  Then use Replicator to replicate the filesystems to the new filesystems.  Essentially you are doing a data migration.  You will eventually have to cut the clients over to the new filesystems which creating new shares.  I would create a separate CIFS server so that you can keep the share names the same.  When you do your final cut-over, it will require that clients be disconnected.  While they are disconnected, you run your final incremental to ensure that the last writes make it over.  Then, when the replicator sessions are broken (and now r/w), you can reconnect your clients to the new cifs server.  You may be able to play some games with DNS, etc to keep the names the same so that clients paths will be the same.

Now that the clients are cut over to the new filesystems, you can delete the old versions of the filesystems on the original file pool.  At that point, the LUNs can be reclaimed by the SPs and the disks in them re-used.

You may want to run some tests before doing this in production.  Or you can always get EMC Professional Services to help your accomplish this as well.

Hope this helps. 

i don't understand why do you recommend that he changes the CIFS server ? If he is using Replicator to replicate vdm/file system he can keep the same CIFS server (i would actually use server_archive for this purpose instead of replicator)

No, we're not  using virtual datamovers. 

you should see your file system using this command:

server_df server_2  ( i am assuming you are using server_2 as your datamover)

do you see your file systems ?

I think we'll have to use server_archive, and selectively move files across, since we don't have quite enough space to fully replicate the entire file system.

I'm pretty much a novice when it comes to the CLI.  I can connect through SSH, but I'm not sure how to find the correct path for my filesystems.  Is there a default setup, or a config file I can view somewhere?

Thanks,

are you using virtual datamovers ?

nas_server -l -vdm

here is an example for you

server_archive server_2 /fs01/dir1/    /fs04/ -rw -J

in this example i am going to copy content of folder dir1 to root of file system /fs04. Please note that dir1 will not be created on fs04, if you want to copy content of dir1 to dir1 on fs04 ..you will need to manually create dir1 on fs04, set matching ACLs and then run the command.  "-J" parameter will preserve ACLs on child directories and files.

Awesome, thank you so much for your help!

Turns out I was pretty close to begin with, I just forgot that the directory paths in Linux are case-sensitive ><