I just added a CX-2PDAE0-FD DAE to our old Celerra NS20. I used USM to add the disk shelf and as we were running the Celerra at 2 Gbps (instead of 4 Gbps), everything went fine.
Once the DAE was added, I used Unisphere to connect to the Clariion (CX3-10) side of the Celerra and configured a RAID 5 (4+1) RG on 5 x 300 GB FC drives. I then created two Luns of the same size on different SP’s and then added them to the NS20 Storage Group. All normal so far.
When I recan the Celerra, I get the following:

17716810659: server_2 c16t1l6 skipping unmarked disk with health check error,
CK200074600886 stor_dev=0x0016, RAID5(-1+1), doesn’t match any storage profile

After a bit of fruitless searching I did the following to fix:

I rebooted both Storage processors on the Clariion side first. (One at a time obviously).
I then rebooted the Standby Data Mover and then suddenly the Celerra could see the DAE and the disks properly. I did not reboot the master Datamover (although I will be doing that just for good measure).