Video game studio leverages Cloud Spanner to improve database performance, with help from Pythian

Customer Success

Video game studio leverages Cloud Spanner to improve database performance

A video game studio needed to modernize their backend infrastructure stack on the cloud to scale with ease and optimize costs. Pythian created a proof-of-concept (PoC) to evaluate the feasibility of Google Cloud Spanner.

Overview

A leading video game design studio was looking to modernize its backend infrastructure stack and game database, hosted in on-premise Microsoft SQL Server Enterprise Edition, to allow for more distributed computing, easy scalability, better performance, and optimized costs. The client’s existing SQL Server databases, combined with an intermediate data access layer in C++ for database communications and object marshalling, were not able to scale up and down quickly enough—leading to frustrating bottlenecks as players were forced to queue to log in and wait for matches.  

Since on-premise game databases are often extremely complex and can be a challenge to scale and manage, even when well optimized, the client wanted to evaluate the feasibility of Google Cloud Spanner by partnering on a PoC with Pythian. 

With expertise in multiple database technologies, along with 20-plus years of on-prem, hybrid, and cloud data management experience, Pythian had the ability to quickly understand the unique complexities and requirements of the client’s source and target databases. 

What we did

  • Developed a PoC to evaluate Cloud Spanner’s feasibility and performance under load, the ease of migration, the impact to existing architecture and design, and to identify any risks or issues
  • Established a migration approach involving a sample of 65 database tables moved from SQL Server to Cloud Spanner
  • Analyzed the client’s SQL Server environment to identify which items needed technical workarounds or refactoring to provide similar functionality on Cloud Spanner
  • Deployed Apache NiFi on a Google Cloud VM to extract data from the client’s SQL server
  • Developed a helper script to streamline implementation by creating a pipeline for SQL Server objects, with the data staged in AVRO format on Google Cloud
  • Deployed Google infrastructure—including Cloud Spanner, Google Compute Engine for NIFI, Google Compute Engine for SQL, and Google Compute Firewall for Resources—through Terraform, which allows the deployment of infrastructure as code
  • Recommended Cloud Identity and Access Management for security purposes, as it allows admins to grant permission to users or groups without modifying Cloud Spanner instances or individual database permissions
  • Educated the client on how to optimize the performance of Cloud Spanner instances 

Technologies used

  • Google Cloud Spanner
  • Google Compute Engine
  • Google Compute Firewall 
  • Cloud Identity & Access Management
  • Stackdriver, Apache NiFi
  • Apache JMeter, Terraform

Key Outcomes

By modernizing their backend infrastructure stack on the cloud, the client was able to reap the benefits of distributed computing across different geographical regions, along with flexible scaling as they continue to add new games to their portfolio. 



Improved database performance 
with a cloud migration strategy, PoC development, and implementation


Explore our Database Services 

No matter your business, no matter the challenge: Pythian’s solutions drive results. 

More customer success stories