Relational databases are awesome. They support foreign keys, transactions and SQL. They are powerful, well-understood, well-supported and battle hardened. But because they support the powerful SQL, it is easy for an unexperienced developer to write a query that brings the whole database down. They also do not scale beyond a single node.
What if there was a way to keep all the advantages of a relational database, but have the ability to horizontally scale them like a NoSQL system? What if there was a system that provided guardrails that protected databases from toxic queries? What if all this could be achieved with minimal changes to your application while letting it view your sharded MySQL cluster as a giant monolithic database?
The year was 2010 and two early engineers at Youtube were contemplating these problems. They built the open source system Vitess to solve them. Over the years, as Youtube databases moved from proprietary hardware to Google’s internal cloud (Borg), it acquired all the features a distributed system needs to work well with ephemeral instances in cloud - excellent observability, robust service discovery, and a solid master failover story. In 2015, Vitess was discovered by companies outside Youtube and now companies such as Slack, Square (Cash), JD.com depend on Vitess for scaling and serving their relational data.
At PlanetScale, we help businesses solve their relational data scaling problems by deploying and operationalizing Vitess clusters on premise as well as in cloud. We provide PlanetScale Deployment Platform that automates the deployment and management of Vitess clusters. We also plan to build a cloud-agnostic (AWS, GCP, Azure) DbaaS solution based on Vitess.
You know how to run a reliable service. You know how requests get from a user to you and back again and you have optimized each step. Your systems don't go down, other people's systems go down and you know how to respond when they do. You are obsessive about data, you assume things that aren't measured have already failed. You fight for milliseconds and live to solve for the 99th percentile. You dream of building high availability distributed services that scale and scale and scale.
What You Will Be Doing
PlanetScale’s goal is to make it a breeze to deploy and operate Vitess/MySQL clusters on premise as well as in cloud. Your job is to make sure our infrastructure and architecture scales to meet those and many more demands, especially as we support multiple clouds, evolve the platform, and run our own database as a service based on Vitess.
- Building out our SaaS Offering on AWS, GCP, Azure and Behind the Firewall
- Supporting our early customers and creating the processes to make them successful
- Working with the Open Source Vitess community to make running Vitess as simple as possible
- Upgrading the PlanetScale Deployment Platform for easily provisioning and operating Vitess/MySQL clusters
What Will Help You Be Successful
- Pragmatic approach towards tackling unforeseen challenges
- Experience with Golang
- Strong Networking Background
- Experience developing and or operating system in public clouds
- Experience with MySQL and SQL
- Infinite appetite for learning
Perks At PlanetScale
- Work closely with a cross functional team of highly motivated and intelligent folks with a unique range of startup and enterprise experience
- Vibrant company culture with frequent team building events
- 20% time for working on projects at company not related to your expertise
- Competitive salary with stock options
- Benefits & 401k plan option
- Fully paid parking pass or CalTrain pass
- Unlimited in-house snacks & drinks
- Unlimited vacation
Founded in 2018, PlanetScale helps businesses solve their relational data scaling problems by deploying and operationalizing Vitess clusters on premise as well as in cloud. The company is headquartered in Mountain View, CA, and is backed by SignalFire. PlanetScale has an experienced team from innovative companies like Paypal, Google, Youtube, Dropbox and SevOne. To learn more about PlanetScale, visit www.PlanetScale.com.