You're already using CEPH-based storage cluster, and would like to analyze and optimize your systems professionally? Or consider the use of Software Defined Storage solutions with CEPH to offer your own highly scalable storage and cloud solutions, and seek expert advice / consulting?

Then you are at the right place: it concepts Oliver Liebel - your Expert/Specialist for HA-Storage-Cluster-Systems and Software Defined Storage.

Whats this Thing about Ceph? - A short Story around a large Scalable System...

Traditional Storage Cluster already encounter their limits in terms of scalability and flexibility.
The rapidly increasing data growth in almost all areas of IT is far beyond scope of the scalability of traditional hardware-based storage.  And - No end so far: All operating worldwide, renowned market research institutes assume that global data volume will reach more than 40,000 exabytes(!) up to the Year 2020.

Regarding studies of the IDC, the majority of companies is not prepared to manage this flood of data. All of these Companies have to deal with rapidly growing data volumes. If they can't, they'll have to handle significant financial losses.

Unlike proprietary, hardware-based storage solutions, CEPH provides an open, software-defined storage platform ( "Software Defined Storage"), which allows almost unlimited scalability of Storage with physical, virtual and cloud-based resources. It provides tremendous cost savings, avoiding the typical vendor lock-in, and also allows unlimited capacity expansions.

The open source, scale-out storage solution CEPH blurs the boundaries of "stale" Storage far behind, and will give you 3 ways do deliver your Storage to your Business or to your Customers:

- CephFS: the highly scalable, Self Healing Cluster-Filesystem for direct Use

- RBD: The Rados Block Device - multi purpose high Scalable and Self-Healing Blockdevice, that can act, e.g as a Backend Device for general Storage Purposes, or as a direct KVM/libvirt VM-Backend, without the need of a Filesystem (see below)

- Rados Object Store / Rados Gateway:  Build your own, massively Scalable Cloud Storage,  directly on top of your Ceph-Scale Out Storage Cluster Pools.

CEPH clusters can be expanded with new nodes and/or Disks at runtime, no special hardware is required, Geo-Site Cluster between data center are daily business, and to make the best and most effective Use of your Storage Hardware, Ceph has Hot-/Cold-Storage policy-based Features already builtin and implemented, to reduce the Costs per TB, and give you the Redundancy you need.

CRUSH-Maps (Controlled Replication under Scalable Hashing) gives you the greatest flexibility, to adjust the Ceph-Storage Layout according to your real Hardware Layout: e.g. From Clients view, Data-Chunks are splitted on datacenter Level, then down to room-level, rack, nodes, Disks - you've got full control down to the smallest chunk of Data.

Ceph-clusters are Self-Healing  - Data on broken OSDs (the Ceph Disks/"Workhorses")  will automatically rebuild on other OSDs. 

Deployment of virtual machines raises a new Level: Through RADOS Block Device (RBD), virtual machines can be deployed (no need for a cluster FS for Live Migration) directly on a RBD-Image. With RADOS Gateway, CEPH Cluster can provide S3 / Swift compatible cloud storage, and traditional claims to a cluster FS are handled with CephFS.

Moreover, NFS / iSCSI on top of CEPH / RBD can be perfecty combinded with the Reliability and Intelligence of Industry proven Pacemaker Cluster Systems, to ensure maximum HA to your Storage Based Applications.

CEPH, by far the most scalable Software Defined Storage solution of the World, is already in Use by many Enterprise Scale production environments, with high reliability and great success.

Almost all Linux Enterprise Cloud Solutions (SUSE / Red Hat) are now based on CEPH.

[click here to get back]