SHARE

Ceph Advances Open Source Distributed Filesystem to the Cloud

Cloud computing presents a new model for data delivery and data storage and it’s a new model that is set to soon benefit from the Ceph filesystem. The open source Ceph filesystem is a distributed filesystem that is intended to be massively scalable. The system was initially created by Sage Weil, who is also the […]

Written By

SMK

Sean Michael Kerner

Aug 1, 2011

Datamation content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More

Cloud computing presents a new model for data delivery and data storage and it’s a new model that is set to soon benefit from the Ceph filesystem.

The open source Ceph filesystem is a distributed filesystem that is intended to be massively scalable. The system was initially created by Sage Weil, who is also the co-founder of hosting provider Dreamhost. Dreamhost is now in the process of building out hosted cloud computing and storage products that will leverage Ceph. Dreamhost is also working on integrating Ceph with the OpenStack cloud computing framework.

Ben Cherian, general manager of emerging businesses at DreamHost, explained to InternetNews.com that Ceph solves a number of problems that other filesystems do not. He noted that Ceph founder Sage Weil found that the metadata layer in most filesystems is not scalable enough. Weil examined popular open source filesystems including GFS and Lustre as part of his research.

Ceph became an open source project in 2006, and currently has 7 full time developers working on it at Dreamhost. Ceph became part of the mainline Linux kernel in May of 2010 with the 2.6.34 release.

Cherian explained that Ceph is a distributed storage platform with three large tiers on the platform. One tier is the distributed filesystem that is POSIX compliant; another tier is distributed block storage. So if a user is access it from a Linux machine or a virtualized environment, storage can be accessed as though it was a block device. The other element is an object store that can be accessed via a RESTful interface.

While Ceph is a solution that was founded by the founder of Dreamhost, DreamHost hasn’t yet entirely moved their shared hosting infrastructure to use Ceph at this point.

“Internally, we haven’t moved over our storage demands to Ceph just yet,” Cherian said. “But that’s just because of migration of a huge platform takes a lot of time and testing.”

Though Dreamhost is not currently deploying Ceph, that doesn’t mean they don’t fully believe in the project.

“In a couple of months we will be releasing a hosted object storage product built on top of Ceph,” Cherian said. “So if our user wanted to use Amazon S3 type storage we’ll have a platform that we’ll be able to offer.”

Cherian noted that the Dreamhost Ceph storage platform should be out in the market in the next few months.

“We’ve been working on Ceph for 8 years with no revenues attached to it,” Cherian said. “We really do believe in the idea that storage should be scalable and storage is the basis of everything that’s on the Internet.”

Sean Michael Kerner is a senior editor at InternetNews.com, the news service of Internet.com, the network for technology professionals.