Evaluating Ceph Deployments with Rook

Ceph is an open source distributed storage system that is designed for excellent reliability, performance and scalability. It was founded in 2006 and since then it has been adopted by a large number of tech giants such as RedHat, Open Suse, CERN, Cisco,
SanDisk, Fujitsu etc. One of the main reason for its popularity is it provides its user with object storage, block storage and file storage all in one. It is a de facto persistent storage for block devices used by OpenStack and hence it becomes a first choice for vendors using OpenStack. At CERN, in addition to being large-scale Ceph users, people at CERN are also active contributors to Ceph components. CERN has the largest cluster of Ceph deployed in its data center. It is used for several storage use-cases such as, for:

• OpenStack Images and Volumes (RBD)

• HPC scratch spaces (CephFS)

• Private NFS-like file shares (CephFS)

• Object storage compatible with Amazon S3 (RGW)

CERN has to deal with petabytes of data so it is always on the look out for ways to simplify its cloud-based deployments. It has been actively evaluating container-based approaches that build upon its Kubernetes infrastructure. One such technology that re-
cently caught attention was Rook; a storage orchestrator for cloud-native environments.