Don’t Forget the Data: the Value of Ceph Storage in Rackspace Private Cloud Powered by Red Hat

cloud storage

One of the most important assets in any business is its data.

Without data, operations are stalled and business is at risk for everything from financial loss to potential loss of life. Yet amazingly, data is often entirely forgotten when building a private cloud.

Dependable shared storage is critical when building private clouds. It not only provides a store of glance machine images to deploy instances, but more importantly, acts as the foundation of the running cloud, with active instance storage held on the storage system.

Ceph Storage

To answer this need, Rackspace Private Cloud powered by Red Hat comes with RedHat Ceph Storage, a scalable and reliable software defined storage solution which meets the storage foundation requirements of an OpenStack private cloud.

RedHat Ceph Storage also offers segment leading self-management and self-healing capabilities, enabling cloud operators to rest easy at night, as sufficient hardware, disk and node failures are handled automatically and seamlessly with no impact to operations. Failed components can be replaced in scheduled maintenance windows rather than in frequent operations required almost immediately to maintain legacy storage system reliability.

Flexibility is key in a private cloud environment, where time to respond to growth spikes and new workloads can be short. RedHat Ceph Storage has the ability to add additional storage nodes that can bring both improved performance and capacity in an online and non-disruptive process, again with no impact on end user operations.

Business applications tend to have long life-cycles, far outliving the infrastructure they’re deployed on. Realistically, that means critical business applications are not designed, and so are not out-of-the-box ready for true cloud platforms. That means that a middle-ground is needed, to answer the requirement of infrastructure resiliency to enable the reliable operation of those legacy applications. Ceph’s scale-out architecture with no single point of failure provides the reliable storage infrastructure those applications require to exist in modern clouds, and enables the live migration of running instances between compute nodes without affecting production operations.

One of the best features of RedHat Ceph Storage is its ability to boot from volume, which allows for speedy instance boot times, by utilizing copy on write clone technology — where a thin clone is created for each instance. This reduces the amount of data that needs to be moved across the network to the compute node, and reduces the amount of capacity consumed on the storage system itself. Faster instance boot times lead to less lag in responding to user demand when scaling out. In some environments, the creation of an instance could be as heavyweight as copying the entire image to the compute node and then back to another volume on the same storage system.

Not all data is created with the same business value, and that data’s value can change over time. Handily, a Ceph cluster can be configured to implement storage tiering, which can help move data to the most appropriate storage tier based on access patterns. That allows the Ceph cluster to be high capacity and cost effective by only keeping a subset of data in SSD storage.

Not all data are created with the same business value. Data which is frequently accessed and needs to be immediately available can be placed on performance optimized storage, such as Solid State Drives. Data less frequently accessed can be placed in capacity optimized storage such as Serial Attached SCSI drives. A Ceph Storage cluster can be configured to use these different tiers of storage within the same cluster so data can be placed on the appropriate storage that matches business requirements.

RedHat Ceph Storage increases performance in other ways. Cluster nodes can be tuned for performance requirements with SSDs, and nodes tuned to capacity with SAS or SATA disks. Ceph also enables multi-protocol access to storage, be that Block for Cinder and Nova-Boot via RADOS Block Device, File, or Object via Rados Gateway as a drop-in replacement for OpenStack Swift or AWS S3, without having to resort to multiple storage systems to provide each of these services.

In short, RedHat Ceph Storage offers an answer to multiple storage requirements and the all-important issue of storing and protecting our data.

And while RedHat Ceph Storage is a powerful storage solution for OpenStack private clouds, many customers prefer not to take on that management, preferring to hand that task to an expert. That’s where Rackspace comes in, with our managed private cloud offering, based on Red Hat OpenStack and Red Hat Ceph Storage.

Want to learn more? Sign up today for a free strategy session. Rackspace has successfully scaled OpenStack to thousands of nodes and operate OpenStack clouds for some of the largest companies in the world. Talk to our proven experts about your business objectives and we’ll strategize with you the solutions needed to achieve them — no strings attached.

Philip Williams is responsible for the integration of enterprise hardware and open source software defined storage with Rackspace's Openstack Private Cloud products. He joined Rackspace in 2011 after stints at Yahoo! and EMC in storage-related roles. He has designed solutions and infrastructure to support many different industry sectors including Rackspace’s multi-tenant environments, massive scale internet properties and financial institutions. Philip holds a BSc (hons) degree in Computing Science from UMIST, Manchester, England, and now lives with his wife in Austin, Texas.


Please enter your comment!
Please enter your name here