May 2018 – Present · 8 yrs 1 mo · Greater Seattle Area

• Tech Lead of Data Pipeline Team
• Owner of CDC Event Bus service, a scalable and configurable system to stream online databases(mongodb, dynamodb) changes to offline databases for analytic processing.
• Owner of MongoDB. Lyft hosts several the largest mongo clusters in the industry. I am responsible for performance tuning, design review and tools/libraries development.
• Major contributor of Database Migration project, a huge data (multiple billion of docs) migration system. The builder of live-oplog and historical-data backfill pipeline, which is an essential component to make sure no downtime migration.

Sep 2008 – May 2018 · 9 yrs 8 mos · Greater Boston

• 2016 – 2018.5
• Leading design and implementation of an IAAS (infrastructure as a Service), which is hosting ECS (Elastic Cloud Storage) for our strategic customers.
• Designed Geo-Pinning consistent hash cache for Global Load Balancer, to distribute Object CRUD traffic evenly to multiple sites. Greatly reduced request latency and recognized by customers and management team.
• Led initiatives to define, design and implement DevOps solutions to host the cloud-based system, includes Load Balancing, High Availability, Monitoring, and Security.
• Containerized all core components and services with lightweight containers (Docker) and replaced previous Virtual Machine based solution. This initiative reduced 20% hardware expense. Got Silver award company-wide for the outstanding contribution to this project.
• Design review of various components in Cloud Storage, includes distributed file system layer and key-value database layer with SSTable and LSM (log-structured Merge Tree).
• 2013 – 2016
• Technical lead of Security and Infrastructure team of CoprHD -- software-defined storage platform for building a storage cloud.
• Designed and implemented Authentication and Authorization layer to support multiple tenant management.
• Architected and implemented IPSec solution to secure communications among nodes in federated data centers.
• Lead efforts in diagnosing and optimizing performance issues.
• Worked on configuration and data management layer using NoSQL database Cassandra and Zookeeper.
• Designed and Implemented Stress Test Tool based on model-based-testing concept
• https://coprhd.github.io/
• 2008 – 2013
• Technical lead on Atmos (EMC Object Storage System) Automation team
• Developed test framework named MaaSClient single-handed, which made test the product’s REST APIs became very easy. Dev team valued the framework and used it as a dev self-testing tool.
• Developed the initial version of performance tool for ECS with grinder.

Apr 2004 – Aug 2008 · 4 yrs 4 mos · Beijing, China

• Worked in Device Drive team of Operating Platform Group
• The major contributor to developing a Series Attached SCSI HBA device driver for Solaris.
• Took the following internal training in Sun Microsystems: Solaris Internal, Writing Device Driver, Network Internal

May 2001 – Apr 2004 · 2 yrs 11 mos

• project involved: Radius Server development, Network Billing System development.