Unsolved
This post is more than 5 years old
2 Posts
0
1301
November 11th, 2015 08:00
Any issues using an RDM for SourceOne Storage?
The guy that implemented our SourceOne had just been using 2 TB vmdks attached to the server and when one would run out of space would have to point SourceOne to a new location. I'm thinking of creating an RDM and moving everything to it. Looks like max size I can make is 16 TB and I currently have about 8 TB of space used and we've been going on SourceOne for at least 3 years so that should get me about 3 more years I hope. Any reason I should not use this method?
Any other recommendations for storage for SourceOne?
No Events found!
bbartel78
2 Posts
0
November 11th, 2015 08:00
Just for clarification 16 TB is the max size I seem to be able to make a LUN in my VNX, not the limitation of RDM.
TheMigrator
25 Posts
0
November 12th, 2015 07:00
To answer your question in a broad general sense, yes you can consolidate the stores. However look at how that usage is now, look at how those VMDK's are spread across the physical storage on the back-end, and pay attention to how the logical use is intended.
With any archive, you of course must pay attention to the I/O capacity and throughput. In general as long as your storage architecture can handle to volume of data through the channels assigned and available, consolidating storage volumes does not itself pose a problem.
Do pay attention to how those volumes are used by the archive, and generally best practices would lead you to make sure that you still want to keep stores separated based on use cases (i.e. historical user, Legal/Discovery Hold, Journaling,....).
When you move those archive volumes around (preferably do this as a copy - Robocopy works quite well) you will need to remap share names appropriately, or you will need to follow up with EMC support or a certified SourceOne partner to help you make sure share names are aligned in SourceOne to their new locations (this is a utility that EMC support monitors very carefully).
When you are doing the realignment of storage resources, do plan on having outage and down time to make the moves and you will need to make sure volumes are properly closed and not being used (pause archive activities and services is the best approach).
If you need further help, let me know and we can discuss, or reach out to EMC support and they can also help you.
RKatwal
2 Intern
•
600 Posts
0
February 8th, 2016 14:00
Greetings,
I would suggest opening service request with EMC support and let them have a look at configuration in place.
Basically it is not recommended to have all things at same place because there are parts which are high I/O intensive, you want to keep those separate.
Data is mainly located at following location:
1. Message Center shares – it’s like a buffer area but needs to perform well in terms of I/O because data goes to this location first for archival and indexing.
2. Index Shares - where indexes resides. You can have many index shares. If one fills up add another one.
3. Volumes location (EMX FILES) – Depending upon what kind of storage you use. This can only point to one location. Unlike index location multiple shares cannot be added.
Regards,
Rajan