By now, we all know that network-attached storage (NAS) provides benefits to organizations of all sizes. However, managing more than a few NAS devices quickly becomes a burden, swapping the value out of NAS simplicity.
NAS consolidation benefits begin with reduced administration points. Eliminating devices that require administration can save time, complexity and money. Good NAS consolidation strategies should take advantage of the declining cost of storage, combine free space and reduce power consumption with low power drives and processors.
Combining devices will create an opportunity to take advantage of the latest technology like clustering, 10 GB Ethernet, asynchronous replication, space optimized snapshots, improved backups and increased data storage capacity.
Many manufacturers of modular NAS filers allow users to upgrade the aging controllers, preserving the data and investment in existing disk drives. This serves to expand the capacity and performance and may allow you to collapse other smaller file servers into the newly upgraded box.
Buying a new NAS server will force you to decide between two broad types of NAS devices. NAS gateways allow customers to leverage their existing investment in SAN storage while integrated NAS products include captive storage. Integrated NAS products install quickly but have limited flexibility and expandability as compared to gateways.
When the existing file servers have significant book value, or the data on them is impossible to migrate, it may be time to consider NAS aggregation. This concept uses a network service or hardware device that proxies NAS requests on behalf of multiple NAS servers. Multiple NAS devices can then function in one namespace, improving flexibility and sometimes simplifying administration. NAS aggregation can be accomplished with products like Microsoft DFS, EMC's Rainfinity and Brocade's StorageX.
Once you've selected a consolidated NAS platform, the hard work of migrating the file server services and the data that goes with it begins. The distributed nature of the NAS clients means that data migrations are never easy. It's usually best to try not to change the access path that end users use to access their data.
The UNC path for a windows file (ex. \\OLDNAS\share\data.file) isn't just used while mapping a drive; it's often imbedded in office documents, scripts and applications. While changing the UNC path in a login script after a copying data sounds easy, don't naively think that you haven't disrupted the user's ability to access their data. It's usually best to try and preserve existing UNC paths, which will require a little extra work on the storage administrator's part.
Whether you're upgrading, consolidating or deploying new NAS, consolidating the general approach to migration will be the same:
- Create a temporary server name on the consolidated NAS device that will be used for migrations (ex. \\NEWNAS\)
- For each share on the existing file servers (ex. \\OLDNAS\share\), create a share on the new NAS server with the same name (ex. \\NEWNAS\share\)
- Copy the data from the old NAS server (\\OLDNAS\share\) to new NAS server (\\NEWNAS\share) while users are still accessing the old NAS file server
- Shut down old file servers
- On the new NAS device, create a virtual server name to match the old file server name (\\OLDNAS\)
- Confirm users can access shares with the old UNC path: \\OLDNAS\share\
If you're deploying a NAS aggregation product, preserving the data path means moving the namespace, not the data to the aggregation device:
- Rename old file server (\\OLDNAS\) to a new name (\\VIRTNAS\)
- Create the old NAS server name in aggregation device
- In the virtualization device, map the old NAS server shares to the aggregation device, presenting the shares to the users
- Ensure users can access data via the old UNC path: \\OLDNAS\share\
While NAS consolidation isn't simple, it's definitely worth doing. There are a lot of options when deciding if you should consolidate, to what you should consolidate and how to do the work. This advice in this tip should help guide you through some of those challenges.
About the author: Brian Peterson is an independent IT infrastructure analyst. He has a deep background in enterprise storage and open systems computing platforms. He has consulted with hundreds of enterprise customers who struggled with the challenges of disaster recovery, scalability, technology refreshes and controlling costs.
This was first published in June 2008