Public cloud measurement report

Public Cloud Performance Measurement Report 2022 by StorPool Storage

Learn how to benchmark your cloud performance against leading public clouds

Most cloud architects understand that raw storage performance is critical for their service quality. Better application performance results in happier users, lower TCO and improved ROI. Public cloud operators rely on performance as a competitive edge to attract and retain customers, which is even more important and harder to achieve in multi-tenant environments.

In this Public Cloud Performance Measurement Report, StorPool presents a testing methodology for evaluating the storage performance of any cloud, in the right way. This report showcases the performance of the block storage offerings of well-known public clouds – Amazon AWS, Google Cloud, Microsoſt Azure, Linode and OVHcloud and compares them against Katapult, a StorPool-based public cloud. 

The tests were performed between October 2020 and November 2021 by StorPool. All tested systems are in production clusters and part of generally available public cloud offerings, so our results are easily reproducible. The Katapult system is part of a production public cloud, so results on this service are directly comparable to results from the big five public clouds.

About Katapult and the StorPool implementation in Krystal’s infrastructure 

Katapult is a virtual Infrastructure as a Service platform designed for extreme performance and simplicity. The solution is developed by Krystal, one of the largest independent UK web hosting companies. Katapult implements best-of-breed technologies and years of successful expertise in the cloud domain, backed up by an exceptional level of service. Krystal selected StorPool because of its high performance, robust API, unique space-saving features and extremely high level of data protection delivered by its triple data replication.

Read more about Krystal’s use case in the case study

About the benchmarking

We performed 4 types of tests:

1. PGBENCH – a database benchmark, perhaps closest to “application performance”

2. Sysbench/MySQL – a second database benchmark for control of PGBENCH results

3. fio – a set synthetic benchmarks — random reads/writes, sequential reads/writes, latency

measurements

4. rsync – copying files and syncing, simulating rapid deployment and backup workloads

Among these tests, we consider PGBENCH and Sysbench/mysql to be most representative of real-world transactional workloads (including most web applications). The remaining fio and rsync tests are somewhat further away from being representative of real-world applications. They serve as a synthetic measure of the ideal latency or maximum throughput possible with each storage technology.

Download the full Public Cloud Performance Measurement Report and learn how to benchmark your cloud performance against leading public clouds

Conclusions

Storage performance determines application performance to a large degree. VMs with seemingly identical parameters like CPU and memory can have orders of magnitude differences in application performance.

Both IOPS and latency are important for application performance.

End-user applications on StorPool-based public clouds perform measurably better, in some examples 2.5x better than the second-best public cloud offering. For another year in a row, the results reveal that StorPool gives superior performance for large-scale cloud Infrastructures and is the fastest and most reliable data storage platform for demanding workloads.

About StorPool Storage

StorPool Storage is a soſtware provider that develops the most reliable and speedy storage platform on the market. It is the easiest way to convert sets of commercial off-the-shelf servers into primary storage systems for cloud infrastructure. Public and private cloud builders – Managed Services Providers, Hosting Services Providers, Cloud Services Providers, enterprises, and SaaS vendors – use StorPool Storage as the foundation for their clouds.

Leave a Reply

Your email address will not be published. Required fields are marked *