R4DAR for Storage

R4DAR for Storage provides monitoring, analysis and management for the next generation of data center storage services. More specifically it is a testing and benchmarking solution for Ceph clusters. R4DAR for Storage complements Ceph management tools provided by vendors and the Open Source community to help manage and benchmark Ceph clusters across their lifecycle and growth.

Run multiple benchmarks in your clusters easily after every change, upgrade or new feature to maintain high quality of service and performance

Archive and compare tests between clusters and measure performance over time to help decide best tuning parameters and optimizations

Identify bottlenecks and optimization points in your Ceph clusters

Ceph Storage

Ceph storage can be challenging to optimize and expand. R4DAR for Storage helps You understand how changes to the storage cluster affect its performance,by running and then comparing benchmarks after each change it is possible to better understand and optimize the performance of the cluster. This way Ceph storage users can get the most out of their storage investment.
Only with R4DAR for Storage you are able to

• Run multiple benchmarks in your clusters easily after every change, upgrade or new feature to maintain high quality of service and performance

• Archive and compare tests between clusters and measure performance over time to help decide best tuning parameters and optimizations

• Identify bottlenecks and optimization points in your Ceph clusters

One time Consulting or Software Subscription

R4DAR for Storage is available as a one time consulting engagement or as a software subscription. When choosing one time consulting, Datalounges experts run the benchmarks and deliver the benchmarking reports with proposals for optimizations. The Software subscription runs in your own Datacenter as a virtual machine.

Please contact sales@datalounges.com
or leave a message in the chat window for pricing

Flexible and inexpensive solution

Ceph provides a flexible and inexpensive solution to solve the problem of rapidly growing storage demand, but can be challenging to maintain due to its high level of automation and flexible services. With Datalounges’ experience as a long time Ceph user and deployment partner for our customers, it became clear that it is necessary to maintain information about the effects of expanding the Ceph clusters, software version changes and effects of tuning attempts. R4DAR for Storage was created to help Datalounges customers stay on top of Ceph storage clusters’ health and performance by providing an easy way to run periodic tests and benchmarks across Ceph clusters.

Periodically running a set standardized tests against your Ceph storage system will help maintain a well performing storage service for the end users. Also the simple dashboard will tell you the status of the Ceph system in one quick glance. The real value of R4DAR for Storage is however that it
· stores anonymized data only relevant for analyzing performance and trending
· creates user visible reports and overviews of time-based growth and performance degradation, identifying bottlenecks
· enables users do comparison with standard deployments to identify improvements
· makes it possible to compare configurations (ceph.conf thresholds)
R4DAR for Storage is a on premise software installed on a Linux server with a yearly subscription per Ceph cluster to allow usage it’s leading functionality

R4DAR for Storage technology

R4DAR for Storage installs on current up to date Linux servers such as SUSE Linux Enterprise Server, Red Hat Enterprise Linux, CentOS and openSUSE.

Using the benchmark tool is straightforward once it is connected to a Ceph cluster and reports can be read and exported via an easy to use web interface.

All tests Object tests are made against 2 different pools. The following tests are run in order to give a healthy and relative in-depth overview of the Cluster.

  • · Rados Benchmark Write, this is to test that the Object placement is working at a decent rate. Results include Bandwidth, Number of writes and Latency Averages
  • Rados Benchmark Read, this is to test that the Object placement is working at a decent rate. Results include Bandwidth, Number of writes and Latency Averages
  • Direct Volume Access tests via Block Device:
    · IOPs simulating mounted workloads, Read and Write random
  • Throughput simulating mounted workloads, Read and Write random
  • Latency to check the network when working on mounted volumes, Read and Write
    Training and deployment are available as consultancy from Datalounges.

    Are you not curious how the performance of your Ceph system enhances or degrades over time and how Ceph compares to your other storage systems? Or is it not valuable to understand how changes to the cluster change its behavior.

Contact us

Kimi Aaltonen Datalounges Managing Director

Finland
Kim Aaltonen
Managing Director
kim.aaltonen(at)datalounges.com

Pekka Lindqvist Datalounges Sales Director

Finland
Pekka Lindqvist
Sales Director
pekka.lindqvist(at)datalounges.com

Ivar Lyse Datalounges Norway Country manager

Norway
Ivar Lyse

Country Manager
ivar.lyse(at)datalounges.com

Technical contact
Oskari Koivisto

Cloud Architect
oskari.koivisto(at)datalounges.com