Enterprise IT shops coping with the explosive growth of unstructured data need to consider their NAS options and decide whether traditional fixed-capacity NAS devices or the emerging scale-out NAS systems will better meet their file-storage needs.
To help make your decision, here’s a summary of the pros and cons for the two types of NAS systems.
TRADITIONAL NAS: PROS
One of the main attractions of traditional NAS is its simplicity. The systems are easy to install, configure, manage and operate, especially in environments of modest scale. Product upgrades in this “scale-up” category follow the traditional speeds and feeds pattern of replacing a box with faster processors and larger-capacity storage.
Scale-up products are generally mature and have plenty of features and add-on software for data protection, business continuance and storage efficiency. Options include snapshots, one-to-many and many-to-one replication, remote replication and remote snapshots, thin provisioning, deduplication and compression.
Traditional NAS systems can be cost-effective and reliable, particularly for small- to medium-sized businesses (SMBs). They help to consolidate file servers and centralize data protection. They may also be tightly integrated with common business applications and their native management consoles.
TRADITIONAL NAS: CONS
The main downside of traditional NAS is its inability to scale beyond the limits of the system, forcing customers to purchase additional, separately managed boxes when they need to add capacity. Capacity on these NAS boxes may be underutilized if users aren’t able to add capacity because they’ve run out of performance or bandwidth.
As IT shops buy multiple products, NAS sprawl is a common problem, and the stove-piped systems have an adverse impact on floor space, power and cooling, and tend to complicate storage management.
“It becomes a nightmare managing 25, 30, 50, 100, 200 of these individual NAS boxes,” said Arun Taneja, founder and consulting analyst at Taneja Group in Hopkinton, Mass. “Ninety percent of the administrator’s time in typical, traditional NAS environments is moving files from one NAS box to another in order to load balance these boxes.”
SCALE-OUT NAS: PROS
Scale-out NAS systems carry the advantage of scaling capacity and performance on an as-needed basis, far beyond the limits of traditional scale-up NAS. They typically distribute data across many storage controllers, and the systems’ clustered architectures ensure high availability.
Early scale-out NAS gained a reputation as being tough to set up and manage, but the difficulty varies by product. To a large degree, several commercial products have eliminated that disadvantage. Users report that adding a node to an EMC Corp. Isilon system, for instance, can be as easy as pushing a button on the front panel.
Once the system is up and running, scale-out NAS brings the huge advantage of being able to manage and move petabytes of data under a single distributed file system and global namespace, and the systems generally support large volumes. Users, in turn, can maintain floor space requirements, power and cooling costs, and management staff.
NAS options at a glance
NAS options at a glance
Scale-up NAS pros
Simpler to set up and manage initially
More software/data protection functionality
Scale-up NAS cons
Prone to NAS sprawl
Scale-out NAS pros
Capacity and performance scales independently
Scale-out NAS cons
“Whether [the scale-out NAS systems] do that through namespace aggregation or having a true global namespace, as long as the customer doesn’t have any management headaches, that’s what counts,” said Terri McClure, a senior analyst at Enterprise Strategy Group Inc. in Milford, Mass.
Some scale-out NAS systems increase performance and capacity independently, while others require the scaling to take place simultaneously. In addition, some products allow their users to upgrade and service hardware without disrupting client access to data.
SCALE-OUT NAS: CONS
Some scale-out NAS systems carry fairly substantial licensing fees that are tacked onto the incremental costs associated with adding equipment. Plus, scale-out systems tend to lack the feature/functionality of well-established scale-up NAS systems, at least at the moment.
For instance, a scale-out NAS system might offer remote replication, but only the asynchronous variety -- not one-to-many or many-to-one functionality. The vendor also might have optimized the system to handle large files rather than huge numbers of small files and might not offer unified block and file storage capability.
“They’ve got good feature/function for the use cases that they’ve grown up in, but if they’re going to expand into enterprise IT, most systems need to get better at small-file, random-access performance, and they have to expand their feature/function use cases,” McClure said. “They’re getting there.”
Taneja agreed, adding, “It’s only a question of time. It’s happening very quickly right now.”
However, Taneja said current claims by some scale-out NAS vendors that they can scale I/O and throughput linearly to seemingly infinite levels, through the addition of more than 100 or 200 nodes, is “total baloney.”
“You can put them together and they’ll work. But you will have a declining performance curve,” Taneja said. “The curve will be pretty linear for a while, and then at some point in time, it’ll start flattening and, further out, it’s going to actually start going down.”
Ravi Chalaka, vice president of product, solutions and channel marketing in Hitachi Data Systems Corp.’s BlueArc NAS Division, maintained that the future of scale-out NAS is the adoption of standards-based parallel file systems based on NFS 4.1 to provide scalability beyond what current systems can deliver.