When it comes to OpenStack, choosing the right storage for your cloud can be the hardest part of building out your environment. Every storage vendor has its own merits. Some excel in particular use cases, but maybe those use cases don’t match your reasons for deploying an OpenStack cloud. Maybe you already own some perfectly good storage and simply want to repurpose your existing investment. Whatever your situation, choosing the right storage so you get the performance, scale, and reliability you need can be daunting.

 

In reality, object storage vs. block storage for OpenStack doesn’t have to be a choice.  Architecturally, block and object storage are complementary solutions for data center design. SolidFire’s block storage delivers incredible performance and seamless scale. Object storage, such as SwiftStack, provides low-cost, disk-based object storage for archival or deep storage requirements. The decision ceases to be either/or and instead can be about how to best incorporate both storage types to get the most out of your OpenStack cloud.

OpenStack supports both object and block storage through Swift and Cinder, respectively. Each have their strengths.

Object storage vs. block storage for OpenStack cloud
Object storage vs. block storage for OpenStack cloud

Cinder Block

  • Storage for running VM disk volumes on a host

  • Ideal for performance-sensitive applications

  • Equipped to handle high-change content, smaller, random reads/writes, and higher/bursty IO

Swift Object

  • Ideal for low-cost scale-out storage

  • Well-suited for backup, archiving or data retention

  • Equipped to handle more static content, larger, sequential reads/writes, and lower IOPS

Initial small or one-off OpenStack projects may lend themselves to optimized secondary storage, when more comprehensive array features, data optimization, ease of scale and resiliency aren’t as essential. However, mission-critical workloads often need the reliable performance and data services that external Cinder block storage provides. Swift object storage often can complement Cinder block storage, functioning as a backup location for primary data files or longer-term archive.

In fact, SolidFire’s Integrated Backup & Restore functionality delivers snapshot-based backups to any object store or device with a Swift-compatible API. Infrequently accessed or archival data can be moved off SolidFire’s production block storage onto object storage, maximizing the benefit of both storage types.

Interested in learning more about optimizing your OpenStack cloud with block and object storage? Download Getting It Right: OpenStack Private Cloud Storage for a more in-depth guide to picking the right storage for your environment. 

mm

Kelly Boeckman