Book Image

Ceph Cookbook

Book Image

Ceph Cookbook

Overview of this book

Ceph is a unified, distributed storage system designed for excellent performance, reliability, and scalability. This cutting-edge technology has been transforming the storage industry, and is evolving rapidly as a leader in software-defined storage space, extending full support to cloud platforms such as Openstack and Cloudstack, including virtualization platforms. It is the most popular storage backend for Openstack, public, and private clouds, so is the first choice for a storage solution. Ceph is backed by RedHat and is developed by a thriving open source community of individual developers as well as several companies across the globe. This book takes you from a basic knowledge of Ceph to an expert understanding of the most advanced features, walking you through building up a production-grade Ceph storage cluster and helping you develop all the skills you need to plan, deploy, and effectively manage your Ceph cluster. Beginning with the basics, you’ll create a Ceph cluster, followed by block, object, and file storage provisioning. Next, you’ll get a step-by-step tutorial on integrating it with OpenStack and building a Dropbox-like object storage solution. We’ll also take a look at federated architecture and CephFS, and you’ll dive into Calamari and VSM for monitoring the Ceph environment. You’ll develop expert knowledge on troubleshooting and benchmarking your Ceph storage cluster. Finally, you’ll get to grips with the best practices to operate Ceph in a production environment.
Table of Contents (18 chapters)
Ceph Cookbook
Credits
Foreword
About the Author
About the Reviewers
www.PacktPub.com
Preface
Index

Disk performance baseline


The disk performance baseline test will be done in two steps. First, we will measure the performance of a single disk, and after that, we will measure the performance of all the disks connected to one Ceph OSD node simultaneously.

Note

To get realistic results, I am running the benchmarking tests described in this recipe against a Ceph cluster deployed on physical hardware. We can also run these tests on the Ceph cluster, hosted on virtual machine, but we might not get appealing results.

Single disk write performance

To get the disk read and write performance, we will use the dd command with oflag set to direct in order to bypass disk cache for realistic results.

How to do it…

  1. Drop caches:

    # echo 3 > /proc/sys/vm/drop_caches
    
  2. Use dd to write a file named deleteme of the size 10G, filled with zeros /dev/zero as the input file if to the directory where Ceph OSD is mounted, that is, /var/lib/ceph/osd/ceph-0/.

    # dd if=/dev/zero of=/var/lib/ceph/osd/ceph-0/deleteme bs...