Editor’s note: In this blog, we look at how Cloud SQL delivered speed, flexibility, and agility to global pizza company Papa John’s International.
With more than 5,400 locations in 50 countries and territories, Papa John’s International is one of the largest pizza chains in the world. In 2017, the company took a cloud-first strategy and kicked off our journey with Google Cloud. We started with a proof-of-concept (POC) project—a very small-scale integration with SADA—for a test set of data. This early test of Google Cloud brought flexibility and agility to the entire software development life cycle of that project. Today, Papa John’s runs on Cloud SQL, BigQuery, and Cloud Storage, as well as Kubernetes Engine, leveraging the power of data to fuel innovation and differentiation across the business, and Google Maps Platform to easily map customer deliveries. Our loyalty programs, our website and our customer and partner experiences are all powered by data. In addition, we’re also looking to equip stores with the power of real-time information to help them make deliveries on time.
For our team, we’re seeing huge benefits on the database administration side. The scalability has become easier, provisioning is faster and the level of productivity is high. We’ve also significantly reduced our licensing costs, which also improves the bottom line. Now that we’re fully managed by Google Cloud, our DBAs and other teams can focus on new initiatives that grow our business and meet our customers’ needs.
Cloud SQL is at the heart of our data initiatives
Cloud SQL has played a huge role in our data initiatives because it is fully managed, reliable, and secure and allows our developers to focus on value-added tasks. We use Cloud SQL with both MySQL and PostgreSQL, depending on the project. We began our transformation journey with a number of re-architecture projects, including revamping our loyalty program and moving our mapping system onto Google Maps Platform, which includes hours and locations for individual stores. This data was managed by our teams on-premises in Oracle, and we successfully moved it to the managed Cloud SQL for PostgreSQL database on the cloud.
For this re-architecture project, we use Google Maps Platform to locate whether a customer’s address is within a store’s delivery zone.
Another project where Cloud SQL has been instrumental is in our commerce platform, which is built on Google Cloud. We partner with a number of third-party aggregators for deliveries and we need to make sure they’re integrated from both data and communications perspectives.We sometimes call the Customer Price Indexes (CPIs) of these aggregator partners to send them menu data, including product availability in stores, product configurations, and more. Our Java applications read that data and call the CPIs for our partner aggregators. Depending on the partner, integrations either push to the partner or respond to the partner’s request right away. We’re also running a call center for ordering, so our staff need to be able to respond quickly to ensure customers are satisfied. All of this data is stored and served from Cloud SQL.
We have about 15 engineering teams following an agile process that uses our various databases. We have one central platform engineering team that uses Terraform and Google Kubernetes Engine (GKE) to provision databases. Over the years, a huge benefit we’ve seen is on the database administration side. That team doesn’t have to do as much hands-on management as they were doing with the on-premises solutions. We worry less about the number of connections to the databases and how many applications are using the data. And since we are building our new databases on the cloud instead of on Oracle, the potential licensing costs are dramatically lower.
Life in the cloud has brought us immense benefits
Google Cloud has helped transform our business. We use Cloud SQL with analytical systems like BigQuery, Google Cloud’s modern data warehouse, and most of these system’s real-time ingestions are event-driven.
In a couple of use cases, we have multiple subscribers for a queue, and those subscribers would process the data and then -depending on the use case, would pump it into BigQuery for analysis. We have found this is a better approach rather than simply replicating all the data going into Cloud SQL into BigQuery.
Looking ahead, we’re planning on moving our entire data systems from on-premises onto Google Cloud, which will help us reduce the footprint of our current databases and use better and lighter technologies. Our experience so far with Cloud SQL, BigQuery and other services makes us confident we’ll reach these goals and continue innovating with our roadmap.
Compared to life before migration, it’s amazing to run on the cloud. With Google Cloud, the provisioning is so easy, the level of productivity is so high, and we don’t have to wait for all of the related infrastructure to be built up. It’s faster than auto-provisioning through several environments, and the need for troubleshooting is greatly minimized.
Cloud BlogRead More