02-10-2015 03:43 AM
We have Enterprise Vault 10.0.4 and currently archive MS Exchange 2010/2013 Mailboxes and Clustered MS File servers (windows 2003).
My question is how would I go about migrating my existing vault stores and partitions from the current file servers to our brand new Hitachi Content Platform Storage system?
I'm assuming downtime would be required whilst the copying of partitions took place.
02-10-2015 04:55 AM
Welcome to the wonderfull world of HCP.
I assume you use the HDS Storage Streamer API to target the HCP? If so, I believe your only option is to move the archives using either 'move archive' or a 3rd party tool (QuadroTech Archive Shuttle or TransVault).
If you can map a drive-letter, then there is documentation on how to move a Vault Store Partition.
http://www.symantec.com/docs/TECH35742
02-10-2015 08:11 AM
what gertjanA is saying applies whenever you go from traditional storage to object-based or (vice versa)
Ravi, are you located in the US by any chance? we have a strong partnership with HDS and might be able to help.
02-11-2015 02:56 AM
We've just done the same thing and the move archive was not an option, we have too much data and Vault Cache users made this impractical.
We had HDS consulting and did the below:
We did this on EV9.0.5 and after careful planning it was actually very simple.
A point to note (happy to be corrected on this) is that no new data is added to closed partitions but if users delete data it will be removed. We took this into consideration and decided to accept the risk that there may be unreferenced data in the stores but not in the DBs. For reference we had around 500 million items at the time.
02-11-2015 03:15 AM
Many thanks for the reply.
Do you think we can get a confirmation on the statement "That no new data is added to closed partitions". It seems like common sense but then again I've seen drives holding closed partitions run out of space during mass rehydrations.
02-11-2015 03:55 AM
I couldn't get anything at the time. I had discussions with HDS, our Symantec partner and Symantec Support and the best I got was that nothing new is added but the data is changeable. See what the community says or maybe another support ticket.
Another risk on our part but mitigated based on evidence from our environment. We have 3 Stores in a group, 12x2TB drives across the 3, each containing around 55 million files, sharing is enabled across the group. Our backups were not showing any 'significant' increase in volume/file counts, the partitions were closed between 6 months and 3 years. HDS completed the replication for us using an HDS in-house tool and we were satisfied that the data was replicated.
Another gotcha we had was retention categories (we have '0', forever retention set in EV, and the HCP device setting is based on this) and WORM mode. Ensure you set this correctly at partition setup time, we didn't and had a very bad time correcting it (new Streamer partition and replica of the old Streamer partition without device level WORM mode, then some SQL hacks to remove the WORM mode setting on the partition).
Also this type of solution may not be right for you, I'm not recommending anything, just sharing as it worked for us given our circumstances.
02-11-2015 04:02 AM
In regards to:
Do you think we can get a confirmation on the statement "That no new data is added to closed partitions". It seems like common sense but then again I've seen drives holding closed partitions run out of space during mass rehydrations.
The statement is correct in that WHEN ARCHIVING no new data is being added to a closed partition. However, if you have (for instance) Storage Expiry running, data might be removed. Restoring/Exporting many archives will create *.archxxx files where and when necessary.
The 2nd by definition is not 'new' data, as they are copies of existing data on those partitions, and will be cleaned up after a default period.