Ceph RBD is a highly available block file system that is distributed over the network to different nodes. Ceph is extremely robust and capable of infinite scalability. Since Ceph is a distributed file system, it is also referred to as a Ceph Storage Cluster.
Advantages
- infinitely scalable
- Highly available and performant
- Flexible and autonomic management
- Object Store - RADOS (Reliable Autonomic Distributed Object Store)
- No single point of failure - distribution to multiple nodes
- Software-based and open source
Application area
Our Ceph storage consists of SSD hard disks that provide many TB of highly available storage space via server clusters. This is used by FluidWEB for the virtual machines (KVM,qemu,LXC/LXD).
E-commerce customers using e.g. Magento or Shopware, but also business applications with database-intensive requirements, benefit from the enormous performance provided by our Ceph storage. All data is transported redundantly via 10Gbit fiber switches. This allows us to achieve network read and write rates of 1.2GB/second, and up to 3GB/second read/write access within the virtual machines on the cluster.
Our Ceph cluster storage performs > 800000 IOPS. If you see this in relation to a fast SSD disk (90000 IOPS), you have a good idea of how well the Ceph storage cluster performs.