DX for Nas Migration to Rainfinity File Management Appliance
With DX for Nas set to go off support in a year, we are starting to look at moving to the replacement product Rainfinity FMA. Is there anyone else out there that has done this migration? It was mentioned that there might be a migration script to convert the DX stubs to the Rainfinity FMA stubs without having to recall and re-migrate everything. Does this migration script\process really exist? If not, I forsee an ugly, labor intensive migration.
Currently no scripts/processes exist to perform this migration. The only way to accomplish the conversion is to perform "ugly, labor intensive migration"
Rainfinity has supposedly been working on a migration utility for almost two years now, so don't hold your breath. My project would still be on hold if I hadn't done it myself.
I took the "ugly, labor intensive" route, and it only hurt for a little while. Like ripping off duct tape from places on your body we won't discuss. Recalled all stubs one file system at a time, reconfigured the DHSM connections on the Celerra, then restubbed with Rainfinity. My biggest issue was having the disk capacity to hold the re-hydrated file systems during the migration. So, I was forced to do a multi step migration. I found the checklist I created which may be of some help.
1. Create temporary file system on SATA large enough to handle recalled file system.
2. Perform initial copy of data from production FS to TEMPFS. This will automatically recall all data. (several hours)
3. Take production share offline.
4. Perform incremental copy to TEMPFS to catch any changes. (one hour)
5. Perform Rainfinity archiving of TEMPFS to re-stub files based on policy. (two hours)
6. Backup TEMPFS with Networker. (two hours)
7. Delete and recreate former production file system, or simply remove all data. (concurrent with backup)
8. Remove DHSM connection point to DX-NAS. (concurrent with backup)
9. Re-create production file system in VDM with appropriate sizing. (concurrent with backup)
10. Create new DHSM connection pointing to Rainfinity. (concurrent with backup)
11. Restore newly archived data from Networker to production FS. (two hours)
12. Put share back online.
13. Decommission server/file system entry from DX-NAS.
The estimated times were to help me plan. I used Networker to backup / restore the stubbed file system because I felt it was the most efficient way to move the data around without causing the stubs to recall. I think you can also copy files with emcopy without recalling stubs, but at the time Networker was a better solution for me.
Robocopy mostly because that's what I was most familiar with (be careful with the time granularity setting for modified files), but emcopy would be a good choice also.
ptinius
15 Posts
2023
1
Posted November 9th, 2009 13:00
Paul