Twelve years ago, we recognized the trend towards an exponential growth of data. This inspired us to start Caringo in 2005, with the intent to change the economics of storage. Over the past decade, we’ve matured our product line and helped hundreds of customers solve what seemed like insurmountable challenges associated with storing, protecting and accessing rapidly scaling data sets. In the past few years—as the Caringo product matured and the rate of data growth accelerated rapidly, it became clear that it was not just about storing data, but solving the problem of managing data at scale and increasing the strategic value of that data.
While other vendors focused on cheap and deep, at Caringo, we expanded our worldview to include “data management at scale.” As a result, we integrated foundational technologies for organization and access (such as search) and developed new ways of organizing data through metadata while allowing universal access through a broad variety of protocols. We also realized that for object storage to be fully utilized, the ability to deploy it for both legacy interfaces and leverage the new paradigm of RESTful interfaces (HTTP) in a seamless fashion was critical.
As a first step to bringing file and object closer together, we focused on integrating existing NAS infrastructure with object storage—without impacting the end-user or administrator experience. Through FileFly, we allow users to transparently offload their primary NAS storage to scale-out object storage and facilitate search and extraction of value from that data—while eliminating backup and enabling a multi-site, distributed, thin-provisioned NAS infrastructure, all at a quarter of the Total Cost of Ownership (TCO) of legacy primary storage.
The next step for Caringo: provide a global address space that can be read and written from any of our supported protocols (NFSv4, S3, HDFS, and SCSP/HTTP) and allow automatic distribution and unlimited NFSv4 mount points across multiple globally deployed object storage clouds—all capable of sharing the same global name space. True distributed file, accessible anytime, anywhere. Meaning you can store it through HTTP, read it through NFS, store it through NFS, read it through HTTP, and essentially use any combination of supported protocols. Legacy file interface integration with the full power of scale-out object storage. SwarmNFS has done just that. It doesn’t just meet our customers’ requests for a global file namespace with universal access (anywhere, anytime) and automated multi-site distribution, it exceeds their requests. SwarmNFS is the first lightweight file converter to bring the benefits of Swarm scale-out object storage to NFSv4.
While most object storage vendors have a file protocol (NFS or SMB/CIFS), their solutions tend to be gateway products that try to mimic or emulate traditional file systems. They are often based on open source software like FUSE or Samba and require local disk caching that must be protected and, in many instances, lacks high availability (HA). This is a transient technology at best: expensive and complex to deploy, maintain, protect and scale. With SwarmNFS, we provide native NFSv4 throughput and a true universal address space that eliminates all caching and spooling for the most cost- and resource-efficient way to enable economical data distribution, management and universal access at scale for organizations.
If you’ve been following our Caringo blog, you know that there are many benefits to using object storage. And now, SwarmNFS enables all of those benefits to be accessible by any organization directly from their NFSv4-based applications, without any custom development. So, your users reap the benefits of a limitless pool of scale-out storage with built-in data protection, high availability, easy distribution, dynamic search, multi-protocol access and the ability to organize data in a variety of ways (including using traditional hierarchy, buckets, or leveraging metadata using common attributes or properties).
Benefits of SwarmNFS include:
- Improved productivity with data portability between NFSv4, S3, HDFS, and SCSP/HTTP protocols
- 80% less resource usage (compared to gateway and connector solutions) with no single point of failure
- Bringing the organizational power of metadata to files, eliminating the need for metadata databases
- Lower TCO of data distribution at scale with the benefits of Swarm object storage
- Leverage the HA nature of Swarm to eliminate local disk caching and the clustering required to protect the local disk cache
- Rapid elastic scale-out of NFSv4 via VM or your hardware of choice; grow and shrink to meet demand
If you are interested in learning more, join Caringo Product Manager Glen Olsen for the What Your Object Storage Vendor Isn’t Telling You About NFS Support webcast today, August 23, at 10 AM PT/1 PM ET.
Glen will also be presenting The Evolution of File webcast this Thursday, August 25, at 7 AM PT/10 AM ET.
What are the characteristics of each storage tier and when should you use NVMe, SAN, NAS, Cloud, Object, or Tape Storage? More Details »
While object storage isn’t a panacea, it is an increasingly important storage technology that enables on-demand access for video workflows. More Details »