In the past, we have worked with organizations that have applications they would like to natively integrate with object storage—gaining the ability to both input and access their information via HTTP protocol and over the web. As the discussion evolved, the question inevitably would come up, “I still have all these home directories, all these file systems, all these file server silos. Is there any way that I can take advantage of object storage for those environments?”
Up until the September 2015 launch of FileFly, the answer was a resounding “no.” It has been difficult to marry the file system hierarchical structure to the flat object/address space that is provided with object storage. Since Caringo was founded in 2005, the problems around file system storage have become more and more complex and exasperating.
At an event a month ago, I was speaking to a storage administrator of a large university who was just about to go buy another Windows based file server and he said, “My file servers have become expensive data silos.” He went on to say, “I have a number of users and I’ve done directory listings. More than 80% of their files haven’t been touched in over a year. And, yet when I talk to the users they say all of their data is important. If there’s a way that I can offload some of that information onto more economical storage that could save IT budget for use in other, perhaps more strategic areas, that would be very interesting.”
What’s the solution? Caringo FileFly, which offers a seamless way to get the performance of NetApp and Windows file servers and combine that with the simple, bulletproof, limitless scale of Swarm software-defined object storage. Texas-born film director Robert Rodriguez said that “Changing the game is a mindset.” The game changer here is that we created a solution that accomplishes this without changing the way applications or users work, which means you can keep your existing mount points. Check out this video that illustrates how FileFly works!
What are the benefits of this marriage between the file-based storage world and the flat address-based/object-based storage world that is scaled out on the backend? We’ve created transparent file access with native integrations to Windows or NetApp file servers. This provides complete support for SMB versions 1, 2, and 3 as well as NFS.
If 80% of the data on your file servers hasn’t been accessed in quite some time, that data could be a prime candidate for moving transparently into Swarm object storage, without your users even knowing (FileFly will leave a small pointer on the file system itself and move the entire file over to Swarm, so it can easily be retrieved when needed). That gives you 80% of your capacity back on your existing file servers while allowing you to consolidate files from many different file servers into that same single address space. Based on our calculations, we’ve estimated that the cost per gig in savings is about 4X versus having to purchase more primary storage in the form of a NetApp or a Windows file server.
Once that data is migrated over to Swarm, the benefits of our scale-out object storage comes into play. Those files are constantly protected. They’re consolidated in a flat address space. You could search those files. You could say how many zip files do I have? What space is that taking up across my entire organization? Anything that has been moved into Swarm is analyzable via the metadata that is included with them, encapsulated with the object. The data is also searchable with native object-based applications, accessible on the web, your mobile device, etc.
For those of you attending SC15 this year, it is in our hometown, Austin, and we would love to see you. Visit us at booth 489 to learn how to lower storage TCO, improve productivity, and eliminate risk with Caringo Swarm. Come see our demos and talk to our experts! We have a limited number of SC15 expo passes available. If you need one or have any questions, please email us at email@example.com. We also invite you to connect with us on Twitter and LinkedIn.