Deploy MySQL with CSD 3000
Leverage transparent compression in hardware to improve query response times, reduce latency variation, and increase system efficiency.
Key customer benefits
Improve uptime
By using fewer components to meet performance and storage capacity requirements.
Improve user experience
With faster and more consistent query response times for multiple users.
Reduce Total Cost of Ownership (TCO)
Through increased Queries per Node, improved SSD endurance and data storage per GB of Flash, and lower power consumption.
“The merits of using ScaleFlux
Computational Storage Drives are great!”
– Shinya Matsui, GMO GlobalSign Holdings, Public Cloud Group
Challenge: Dealing with scaling demands on the database
MySQL is a popular database deployed to enable high data traffic volumes delivering high performance and low latency with linear scalability. The challenge is that generational improvements in server compute and memory resources are insufficient to meet the increasing demands for database transactions on larger and larger data sets. This forces users to deploy more servers and storage to overcome the CPU and memory bottlenecks. This scaling out of infrastructure only exacerbates other inefficiencies – cost, reliability, performance, uptime, SLAs, and management overhead for IT staff all suffer.
Solution: Leverage ScaleFlux CSDs to mitigate the scaling pains
Choosing a better SSD from ScaleFlux, built with computational storage technology, is the easiest way to improve database performance while reducing the infrastructure required to support MySQL deployments. Using an NVMe SSD that can intelligently compress and manage data inside the drive improves latency consistency, increases application performance, extends drive life, and reduces the amount of hardware needed to meet daily transaction loads – all without additional software installation or app refactoring.
Results: Improved user experience with a simplified infrastructure
In a scenario testing using Sysbench 8.0 with a 1.5TB data set running 64 threads with Double-Write Buffer on and a CSD 3000 installed, the system achieved over twice the queries per second with 2.4x lower 99% latency than the same system did with another vendor’s PCIe 4.0 NVMe SSD of the same capacity.
In the case of GMO Cloud, deploying ScaleFlux CSDs enabled them to dramatically simplify the cluster needed to meet their QPS and storage commitments – cutting the number of servers required by 50% and the number of SSDs necessary by 75% vs. their initial configuration.
Deploy MySQL with 2x the performance on half the infrastructure
Leverage transparent compression.
