NAS vs. Object-Based Storage

Today’s IT professionals basically have two options for dealing with exponential data growth. The first option is to stick with NAS, and commit to continually scaling out the architecture, or they can implement a new approach with object storage technology.

Legacy NAS

For years, IT organizations have relied on file-based storage technologies like file servers and networked attached storage systems to manage PDFs, end-user files, email, and media files. Because it’s fairly simple to deploy and manage, NAS storage has been seen as the path of least resistance when it comes to storing unstructured data.

Recently, however, the limitations of large-scale NAS environments have started to become more apparent. For example, scale-up NAS systems experience performance degradation when the controllers’ I/O limitations are reached. The result is multiple islands of storage, and additional complexity, cost, and risk.

While scale-out NAS systems can support multi-petabyte enterprise systems, they can be expensive to implement and difficult to maintain. Data protection can be a challenge, as well.

While most NAS systems provide data protection tools like point-in-time snapshots, archiving, and replication, getting the full benefit from these features requires a dedicated off-site DR array to safeguard data in the event of a local failure or outage.

The initial investment in a DR system can be expensive in itself, but as data continues to grow the model can become completely unsustainable for scaling out unstructured data.

Object Storage – The New, Old Solution

The solution to cost-effectively storing large amounts of unstructured data may be a technology we’ve known about for years.

For over a decade, object storage systems known as “content address storage” or CAS has been used to store and protect data like medical images, email and corporate documents subject to regulatory compliance mandates. Many of these original platforms were proprietary solutions that bundled together hardware and software and provided only limited scalability.

So while some CAS systems were more cost-effective than NAS, they were unable to scale capacity to handle massive data growth. The result was multiple islands of CAS storage in the data center, and increased complexity and overhead.

The New Generation of Object Storage

The latest generation of object storage technology is designed to solve this issue. With object storage software, businesses can combine commodity storage resources like internal server disk, direct-attached storage, and JBOD (or “just a bunch of disks”) to create multi-petabyte repositories for unstructured data. Some object storage technologies also employ “object dispersal” or erasure coding, a cost-efficient alternative to RAID methodologies.

A potential obstacle for object storage is the fact that existing applications may need to be recoded to work with object storage systems. But if the chosen software platform supports standard protocols like NFS, this issue be avoided. Applications can simply use an NFS mount point to begin storing data without needing costly software programming changes.

Object Storage and the Cloud

Some object storage platforms allows seamless connectivity to cloud storage, such as Amazon’s S3. By leveraging the cloud for DR, businesses can avoid the costs of implementing their own business continuity solution.

While many organizations will still need NAS to support their most resource-intensive applications, the majority of unstructured data can be managed more cost efficiently with vendor agnostic object storage. If your company is exploring storage options, consider managed storage services from Park Place Technologies. Learn more about this option today!