Specialties: DevOps, Golang, Kubernetes, Containers, Automation, Machine Learning Platforms

• Name: Harrison Zhu
• Experience: 14.5 years
• GitHub: Hz89
• Position Sought: Infrastructure Developer / DevOps Engineer / SRE
• Preferred Location: Singapore
• Contact:
  • Email: [email protected]
• Current Role: Employed
• Education: Associate Degree

• Programming Languages: Golang, Python
• Version Control: Git (GitLab, GitHub, Gitea)
• CI/CD: Drone, Jenkins
• Databases: MySQL, TiDB, TiKV, Codis, Redis
• Infrastructure as Code (IaC): Terraform, RKE
• Infrastructure: Kubernetes, OVS/OVN, Flannel, Calico, RoCE, IB, IPoIB
• Container Technologies: RKT, Docker
• Operating Systems: CentOS 6/7, CoreOS, Ubuntu 16.04
• Cloud Platforms: AWS, Alibaba Cloud (Extensive use of Golang SDK)
• Methodologies: DevOps, Agile

• Tencent Cloud EKS:

  • Led architecture and core development for EKS versions 1.0, 1.5, and 2.0.
  • Established CI/CD pipelines and development standards.
  • Developed and managed Windows container support.

• EKS OS Development:

  • Created a customized Linux distribution tailored for EKS, optimizing startup speed.

• CSI on Nodeless Kubernetes:

  • Implemented a unique CSI compatible with nodeless Kubernetes for EKS.

• MOPS K8s Multi-Cluster Management Platform:

  • Enhanced and customized the open-source Wayne project for multi-cluster management.
  • Integrated Alibaba Cloud for DNS automation.

• Jenkins CI Platform:

  • Deployed Jenkins with Kubernetes for build clusters.
  • Implemented caching for Jenkins agents with shared storage.

• Goslurm Deep Learning Scheduling Platform:

  • Developed K8s-based scheduler compatible with Slurm for deep learning tasks.
  • Integrated with frameworks like TensorFlow, PyTorch, and Caffe2.
  • Hybrid cloud deployment with multi-cluster and auto-scaling on AWS.

• Hive Training Dataset Management System:

  • Created a version-controlled dataset management system compatible with AWS S3 and Alibaba Cloud OSS.
  • Provided a Git-like interface for managing dataset versions.

• CacheFS Distributed File System:

  • Designed a distributed, multi-layer caching system for deep learning tasks.
  • Supported various storage backends (AWS S3, Alibaba Cloud OSS, CEPH) and implemented Kubernetes CSI integration.

• UPaaS Private PaaS Platform:

  • Architected a private PaaS including CI/CD, image repository, and elastic computing with Kubernetes.
  • Integrated Gogs for git services, Drone for builds, and Harbor for a private Docker registry.

• Djob Centralized Cron Job Management:

  • Built a high-availability cron job management and command scheduling system using Golang, SWIM, gRPC, etcd, and TiDB.

• CMDB:

  • Designed and began developing a CMDB system for managing configuration and asset information across infrastructure.

• Monitoring System:
  • Implemented a Zabbix-based monitoring system with customized alerts and data storage.
  • Used MySQL-TokuDB and Python scripts to improve data handling for high-volume monitoring.

• TikTok (Singapore)

  • Position: SRE
  • Duration: Dec 2022 – Present
  • Role: As part of the SRE team at TikTok, I focus on identifying and resolving critical stability and performance issues. My role includes monitoring, diagnosing, and troubleshooting various bottlenecks to ensure optimal system performance and reliability.
  • Key Achievements:
    • Identified and resolved IO bottlenecks in the production environment, enhancing overall system responsiveness.
    • Optimized connection pool management in the Go MySQL driver, leading to improved database access performance.
    • Addressed and resolved long-connection issues within the service mesh, improving network reliability and stability.
    • Enhanced the network IO performance of TikTok's internal framework, Kitex, by optimizing configurations and identifying architectural improvements.

• Tencent Cloud

  • Position: Senior Backend Developer
  • Duration: Aug 2019 – Dec 2022
  • Role: Led architecture and core development for Tencent’s EKS platform, focusing on building scalable, nodeless Kubernetes services from the ground up.
  • Key Achievements:
    • Developed and scaled Tencent Cloud’s nodeless Kubernetes to support over 2 million cores, serving 1,000+ enterprise customers.
    • Built a custom Linux OS for EKS, reducing startup time and optimizing system resources for container workloads.
    • Designed and implemented a CSI-compatible storage solution to ensure seamless storage management in a nodeless Kubernetes environment.

• Momenta (AI Autonomous Driving)

  • Position: Development Engineer
  • Duration: Feb 2018 – Aug 2019
  • Role: Supported deep learning platform development, focusing on DevOps, CI/CD, and infrastructure scalability.
  • Key Achievements:
    • Customized the Wayne open-source project for multi-cluster Kubernetes management and integrated it with Alibaba Cloud for automated DNS management.
    • Deployed a Kubernetes-based Jenkins CI environment, reducing build times by implementing caching solutions for shared storage.
    • Developed a distributed file system, CacheFS, for deep learning, supporting AWS S3, Alibaba OSS, and CEPH, and integrated with Kubernetes CSI to streamline dataset access.

• Youzu Interactive

  • Position: Senior DevOps Engineer / DevOps Development Supervisor
  • Duration: Mar 2015 – Jan 2018
  • Role: Spearheaded Youzu’s private PaaS platform, integrating CI/CD, container management, and infrastructure automation.
  • Key Achievements:
    • Architected and launched UPaaS, a private PaaS platform combining Gogs, Drone, and Harbor for code, build, and image management.
    • Created Djob, a centralized cron job management system, achieving high availability and flexibility for task scheduling across Youzu’s infrastructure.
    • Initiated the development of a CMDB system to manage configuration and asset information, enhancing operational insights and control.

• 17173 Beijing Branch

  • Position: DevOps Development Engineer
  • Duration: Sep 2013 – Mar 2015
  • Role: Supported development and operations for a high-performance monitoring solution tailored for large-scale gaming applications.
  • Key Achievements:
    • Developed a Zabbix-based monitoring system with customized alerts and data handling improvements, specifically for gaming applications.
    • Implemented a MySQL-TokuDB backend to manage and store high-volume monitoring data efficiently.

• Shanghai Jiangyou

  • Position: Operations and Maintenance
  • Duration: May 2011 – Feb 2013
  • Role: Provided IT operations support and maintenance, focusing on server uptime, performance, and reliability.
  • Key Achievements:
    • Maintained and supported various applications and infrastructure, ensuring high availability and performance.
    • Implemented regular server and network audits to optimize uptime and resource management.

• The9 Limited

  • Position: Operations and Maintenance Intern
  • Duration: May 2010 – May 2011
  • Role: Assisted with maintenance and operation of The9’s gaming and social websites, ensuring minimal downtime and optimal user experience.
  • Key Achievements:
    • Supported daily operations for NBU1 department’s social websites and gaming platforms.
    • Conducted regular backups, audits, and troubleshooting to prevent and resolve technical issues.

I previously built a nodeless Kubernetes service from scratch at Tencent Cloud, scaling it to support over 2 million cores and serving more than 1,000 enterprise customers. Currently, I am responsible for ensuring the stability of TikTok Shop, a platform composed of over 3,000 microservices.

I am seeking a position with competitive compensation that aligns with my experience and expertise. As an Employment Pass (EP) holder, I would also require visa sponsorship from any future employer.