Essential Guide

Unstructured data storage showdown: Object storage vs scale-out NAS

A comprehensive collection of articles, videos and more, hand-picked by our editors
Manage Learn to apply best practices and optimize your operations.

Nine reasons object-level storage adoption has increased

Scalability, performance and low cost are some of the motivations behind implementing object storage, but there are many more that have organizations looking at the technology.

Object-level storage is seeing higher levels of adoption than ever before. From compliance requirements to scalability...

issues and budget constraints, object storage has proven to be an effective alternative to traditional file storage. Here are nine reasons why more organizations are turning to object storage technology.

Reason 1: Growth of unstructured data

According to IDC, unstructured data has a compound annual growth rate of approximately 62%. The granular approach of object-oriented storage makes it a valuable resource for dealing with large amounts of unstructured data.

Reason 2: Scalability

Object storage can scale into the dozens to hundreds of exabytes of usable capacity, and there is no technical reason why object-level storage cannot scale into zettabytes or even yottabytes. It leverages the highest density storage, whether it's hard disk drives, solid-state drives or both.

Reason 3: No migration required

Object-based storage is a shared nothing scale-out system, so it does not require a technology refresh in the traditional sense. When hardware needs to be refreshed, it is simply a matter of adding the new object storage nodes to the system typically connected with standard Ethernet networks. This is followed by removing the old nodes, one at a time, from the system. No data has to be migrated. Erasure coding or multicopy mirroring, in addition to self-healing algorithms, recreates the data required on the new nodes. As capacity increases, data is automatically leveled across all object storage nodes.

Reason 4: Compatibility with Hadoop and NoSQL

Object-level storage is often a good storage option for Hadoop or NoSQL, depending on the vendor. Some vendors, such as Caringo and Cloudian, even have analytics capabilities built into the object storage itself.

Reason 5: Replaces traditional file storage

Another reason object storage adoption has accelerated is that it is perceived as the next generation of file storage. Most object storage systems available today have NFS, SMB and iSCSI interfaces built into the system. Geographic distribution capabilities and analytics make object storage very useful and valuable as file storage.

Reason 6: More efficient than tape, but still low cost

Next to tape, object storage is the most cost-effective form of data storage. But while tape is too slow for most analytics, active archiving or fast restores from backup, object storage is more than fast enough. Unlike tape, object storage is also very good at sharing unstructured data between geographically dispersed personnel, working with Hadoop infrastructures and keeping the data whole beyond 30 years. Tape and object storage can both implement write once, read many technology, but only object storage can limit access to an object or file.

Reason 7: Improvements to immediate consistency

One of the drawbacks of object-level storage was that it was eventually consistent, which could lead to errors since data was updated concurrently in two different nodes or locations. However, several object storage providers -- Caringo, Cloudian, HGST and Samsung-Joyent, for example -- now offer immediate consistency with object storage technology.

Reason 8: Faster performance with flash SSDs

Object storage was also susceptible to high latency and slower response times. Strategic implementation of much faster, relatively low cost, dense high-capacity flash SSDs based on consumer multilevel cell and 3D triple level cell has been eliminating that weakness.

Reason 9: Standardization

Perhaps the most prominent reason for the acceleration in the growth of object-level storage adoption is the rise of Amazon Web Services (AWS) and Amazon Simple Storage Service. Amazon S3 is AWS' object storage, and the S3 application programming interface has become the de facto standard object storage API. All object storage providers -- even open source variations of OpenStack Swift and Ceph -- have adopted the S3 API at some level to make themselves compatible with applications that already use it. This has made the plethora of applications written for AWS S3 compatible with private object storage implementations.

Next Steps

Is object storage a good fit for the cloud?

How to better protect object-level storage

Object storage startups to know in 2016

This was last published in August 2016

PRO+

Content

Find more PRO+ content and other member only offers, here.

Essential Guide

Unstructured data storage showdown: Object storage vs scale-out NAS

Join the conversation

2 comments

Send me notifications when other members comment.

By submitting you agree to receive email from TechTarget and its partners. If you reside outside of the United States, you consent to having your personal data transferred to and processed in the United States. Privacy

Please create a username to comment.

Why has your organization adopted object storage?
Cancel
Well, a rather breezy article by Mr. Staimer, who is usually "deep diving" when he is expounding on object-based storage. If more organizations are considering object-based storage, I doubt that many of them have exabyte-plus scale deployments in mind.

The problem with gushing about the scale-out capabilities of object-based storage ignores the question about whether object-based storage can also scale down to meet the needs of thousands or even millions of potential users for this storage technology.

IBM Cloud Object Storage (Cleversafe) marketers think 400TB to 500TB of storage is the "comfort zone" for entry level customers. There are probably many more customers "comfortable" with 40TB to 50TB of object-based storage because they don't want to keep upgrading and/or replacing their silos of SAN and NAS storage systems. When you think about object-based storage you need to think about storage simplification.

Some object-based storage vendors, like Cloudian, can scale down to as little as 10TB of usable storage in a cluster. Every object-based storage vendor can scale-out to some ridiculous number of petabytes, but not everyone can scale down to where their storage technology is cost-efficient and easy to use by small and medium-sized organizations.
Cancel

-ADS BY GOOGLE

SearchSolidStateStorage

SearchCloudStorage

SearchDisasterRecovery

SearchDataBackup

Close