Karan Singh

Code Never Lies, Comments Sometime Do !!

Ceph : Object Storage Performance & Sizing Guide

| Comments


To all the Ceph enthusiast out there !!

I am very proud to introducing Ceph Object Storage Reference Architecture : A performance & sizing Bible for deploying object storage at scale

In my current job at Red Hat, i have been working extensively on Ceph object storage benchmarking. During this project we did hundreds of hours of performance testing on a Ceph object storage cluster with a wide range of object sizes and workload patterns. During the course of benchmarking, we learned how to best tune Ceph object storage for your workload.

With more than 1500 [Test ⇒ Tune ⇒ Repeat] exercises and over 50 pages of sheer technical content, this reference architecture provides empirical answers to a wide range of performance questions surrounding Ceph object storage, such as (but not limited to):

  • What are the architectural considerations before designing object storage?
  • What networking is most performant for Ceph object storage?
  • What does performance look like with dedicated vs. co-located Ceph RGWs?
  • How many Ceph RGW nodes do I need?
  • How do I tune object storage performance?
  • What are the recommendations for small/large object workloads?
  • What should I do? I’ve got millions of objects to store.

So, if you are into object storage world OR just about to enter …. this reference architecture is for YOU

Unlock the Performance Secrets of Ceph Object Storage with the help of Red Hat Ceph Object Storage Performance and Sizing Guide.

Shout out to my teammates Kyle Bader & Brent Compton with whom i worked on this project and we are very proud of what we came up with.