Philip Williams
on 6 September 2023
Any data-centric organisation faces the prospect of data growth at some point in their existence; it is estimated that more than 2,500 Petabytes (PB) of new data is created every day. While only a few organisations will ever have to deal with that kind of scale, it is important to plan ahead for your own organisation’s growth.
Legacy appliance-based storage systems where hardware and software are closely coupled can be expensive to maintain, expand, and difficult to operate. Moving between different models of an appliance, or to a more modern generation usually requires a complex fork-lift replacement and lengthy data migration.
Combine these more traditional challenges with the fact that application behaviour is changing too. You can no longer rely solely on a multi-protocol storage system that only provides block and file services. Many modern applications extensively utilise REST-based APIs like S3 and Swift for their storage needs. This is where a software-defined storage solution can help avoid storage silos by servicing all of these different needs from a single cluster.
What is software-defined storage?
A software-defined storage (SDS) system is a type of storage solution that abstracts away the underlying hardware, unlike traditional appliances that are tied very closely to the hardware supplied by the vendor. This means that a user of SDS can use commodity off-the-shelf hardware to build their storage system and still meet their performance and economic objectives.
Taking this concept further by combining commodity hardware with open source software extends this flexibility. Organisations are freed from cumbersome licensing agreements, can increase their security posture by not trusting black-boxes, access truly open standards, and access the latest hardware and software innovations.
Ceph is the leading open source software-defined storage system in the market, and can be used to provide storage to bare-metal servers, OpenStack and kubernetes infrastructure, and even direct to applications via S3 and Swift APIs. For example over 70% of production OpenStack clouds use Ceph as a backend for their block, file and object storage.
Cloud storage for enterprise with Ceph
Ceph is designed to provide massively scalable block, file and object storage from a single highly-resilient cluster. With no single point of failure, support for non-disruptive software updates, and features to allow for local and remote replication, compression and data at rest encryption Ceph is perfect for use in an enterprise setting.
Ceph clusters are built entirely from commodity hardware, either x86 or ARM64, combined with SSDs and Hard Drives. Standard off the shelf Ethernet is used for internal (cluster) and external (client) communication. This use of off-the-shelf hardware leads to a storage system that is not only highly configurable but also economically advantageous when compared with typical appliance based systems. Heterogeneous hardware can be supported as a cluster grows.
As an enterprise’s storage needs grow, Ceph can eliminate storage silos, as a single cluster can expand to meet all storage demands, regardless of the access protocol or performance requirement.
- Block storage needs are provided via the RADOS Block Device (RBD) protocol – a highly scalable multipath-native block transport. To support legacy environments iSCSI can also be accommodated via a gateway, and in a future release NVMeoF will be supported as well.
- Shared File storage is presented either via Ceph’s native POSIX compatible protocol CephFS, or via the NFS protocol.
- Object storage with API compatibility for both the S3 and Swift APIs is fully supported.
Canonical Ceph, a smart choice for open source storage
Over the last few years there have been a number of changes in the ecosystem surrounding Ceph. After SUSE’s merger with Rancher they rapidly retired their Ceph based storage product, and RedHat was acquired by IBM, with their Ceph team later joining the proprietary storage team within IBM.
Throughout these changes, Canonical’s commitment to open source remains unchanged, and with Ubuntu Pro, a Canonical Ceph cluster deployed on Ubuntu today will be supported for a minimum of 5 years, with the option of extending that support for security patches to 10 years.
Where to learn more?
In our white paper, A guide to software-defined storage for enterprises, we take a holistic look at how you can replace legacy appliance-based storage systems with SDS, retain the features that you have become accustomed to, while increasing flexibility and reducing cost.
