Who We Are
Twitter’s storage infra organization (~70 people) is made up of 4 teams - Blobstore, Graph, Manhattan and Cache. We’re currently hiring for Manhattan and Graphstore in EMEA and Blobstore in the US.
Twitter’s globally distributed, real-time communications network generates trillions of events and petabytes of data per day. Manhattan provides the key-value storage layer that enables products and services at Twitter to be fast, highly scalable, and always available.
Graph Storage stores relationships between users or tweets, such as follows, blocks, mutes, favorites, retweets, replies, and so forth. We have ~8 trillion edges and hit 65M QPS each day. Our goal is to enable our customers to think Graph, providing them with storage infrastructure that can scale to their needs.
The Blobstore team stores and serves petabytes of blob data, including the media uploaded by our users. This is critical for Twitter's success, and an opportunity to directly make a positive impact on the experience of every Twitter user.
What You'll Do
Today, Twitter operates in 2 on-premise datacenters. As part of Twitter strategy, we would like to run a cloud datacenter for real-time serving needs. In addition to that, we are focussed on delivering storage on compute on premise (migrating from Apache Mesos to Kubernetes), which will save run-the-business costs, and reduce our overall cost of operation.
You will help us with strategy and delivery of both initiatives, focussing on running storage on Kubernetes and migration to the cloud. You will work closely with Twitter’s compute team.