On Rackspace, via our present monolithic structure, we had been managing giant clusters of cases operating on MySQL. Every of Freedom Monetary Community’s enterprise items had one giant cluster of cases. Rackspace was managing these clusters, and thus taking work off our arms, however we had little or no management over these databases. Each small change corresponding to disk resizing would take a few weeks no less than. Due to that, our database cases had been vastly overprovisioned and costly. 

We noticed that Google Cloud may host and handle all of our databases, saving us invaluable time and sources, and that Google Cloud SQL’s versatility would enable us to construct versatile, safe options that will meet the wants of our groups and our prospects. We had been capable of break down our clusters into many smaller cases that we are able to handle solely via automation with out including overhead.

A posh migration made simpler by Google Cloud

Our migration concerned the transformation of our monolithic structure to a microservices structure, deployed on Google Kubernetes Engine (GKE) and utilizing the Cloud SQL Proxy in a sidecar container sample or the Go proxy library to connect with Cloud SQL. Every microservice makes use of its personal schema and schemas may be grouped in shared cases or be hosted on devoted cases for increased load purposes. 

We efficiently leveraged Google Cloud’s new Database Migration Service (DMS) emigrate our databases from Rackspace to Cloud SQL. We used it emigrate three separate manufacturing databases, with 5 complete schemas migrated and an general dimension of near 1 TB of knowledge with lower than 15 minutes of downtime. In the end, the migration was profitable and largely painless. We’ve shut down our providers at Rackspace, and all of our databases are operating on Google Cloud’s managed providers now. DMS was actually the one possibility due to the dimensions of our databases. We estimated that doing a “dump and load” migration would have required utility downtime in extra of 12 hours—to not point out the hours we might have spent doing prep work. 

Utilizing Cloud SQL as our database basis

Since finishing the migration, Cloud SQL has helped us meet our objectives round safety, scale, and adaptability. We now deploy a strong set of microservices and cases—following a current resizing, we’ve an estimated 180 cases consuming 350 CPUs, for 1300 gigs of RAM. Our microservice examples embrace every little thing from easy use instances and utility configuration databases to bigger, extra complicated databases that maintain info used continuously by enterprise groups. We save a lot time not having to handle 180 cases.

Leave a Reply

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