/ Blog / Size Doesn’t Matter – Here’s Why You Should Cluster

Size Doesn’t Matter – Here’s Why You Should Cluster


Using Ceph for small datasets: Benefits

In our experience, Ceph is the clustering solution. Ceph’s scalability, data security, and availability provide the vast majority of organizations exactly what they are looking for from their storage. At 45Drives, we are the only organization that offers full and flexible support for both Ceph itself and the hardware of the cluster.

Many of our clustering customers come to us because they have a large or rapidly expanding amount of data they need to store, sometimes into the tens of petabytes. But, what about people with a smaller amount of data, say in the range of 5-30 TB? Clustering provides the same inherent architectural benefits and can cost less than a single server would from the big legacy vendors.

What do you get from a Ceph cluster?

High availability: Clustering involves taking multiple data storage servers and linking them into a single storage system. This provides redundancy, even if an entire server goes offline your data will be safe and accessible, which ensures uptime. When using a single server for your data storage, in the unlikely event something goes wrong, such as a component failing, your server is very likely going offline. How much downtime really matters to you is something that is dependent on your organization’s needs and priorities. But, for many organizations, minimizing server downtime is absolutely critical for their success. Using a cluster for as little as 5TB of usable data will still provide enough redundancy to protect against entire servers going down, as long as you set the correct failure domain.

Data security: Another product of the inherent redundancy of Ceph cluster architecture is amazing data security. Replication and erasure coding are both options when using Ceph. Erasure coding is just like RAID except it is implemented at the hard drive level. It can also be applied on the server level, rack level, or any higher levels of abstraction.

Future-Proofing/Scalability: You may only have a few terabytes right now, but what if you need the option for more in the future? Scaling with single servers is limited by drive slots, and it is generally best to avoid the complexity that is introduced by adding different stand-alone storage systems. With Ceph clusters, you can scale easily and infinitely.

With a small set of data in a 45Drives cluster, you can start out with only a drive or two in each unit, then scale for a long time by adding more drives to the empty slots as needed. Drives come up to 18 TB nowadays, and our minimum server size has 15 drive slots. If you fully populate 3 nodes with the maximum number of 18 TB drives, you will be up to 810 TB of raw storage. This can come as a small initial investment, with OpEx purchases covering scaling from that point forward. When you run out of space doing that, you can seamlessly add new nodes to the cluster. Ceph will automatically balance these new additions, and provides a solution that will cover your organizations data storage needs for a long time.  You get this future proofing by choosing Ceph as your solution now, when your data storage needs are still small.

Simple management: Between 45Drives’ expert support and Ceph’s dashboard, it has never been easier or more affordable to store your data. Our support team are experts on Ceph and are available to guide you through the entire deployment and management process, exactly as much as your team requires. We will support any of our customers, whether they are dealing with Petabytes or terabytes — and are completely committed to having your storage production ready.

(Ceph also has a dashboard that simplifies the routine maintenance and monitoring tasks that administrators perform with their storage. Check it out here to get a complete video walkthrough, or here for a written blogpost.)

In conclusion, the necessity to cluster isn’t only determined by how much data you have, but also by whether or not you deem it important to gain the advantages that only clustering can provide.

If you are at all curious about what Ceph can do for your organization, click here to contact us, or here inquire about a free Ceph webinar.

Newsletter Signup

Sign up to be the first to know about new blog posts and other technical resources


The latest news, articles, and resources, sent to your inbox

45 Drives® is a division of Protocase Incorporated. © Copyright Protocase Incorporated. All rights reserved.

Terms & Conditions