San Jose, California, United States
- Deep understanding of Cloud technologies (IaaS) with focus on storage and network. Built a complete private cloud product with proprietary and open source softwares. Technologies involved include KVM/QEMU, OpenStack, Kubernetes, Docker, MariaDB/Galera, InfluxDB, RabbitMQ, Nginx, Helm, Etcd, Openvswitch, IPtables, Canal/Calico/Flannel, VLAN, VxLAN, Ceph storage, Java Spring Boot, REST API, etc - Thorough understanding and hands on experience of network and storage architecture, principles, protocols and products (TCP/IP, DHCP, VRRP, ARP, IPtables, Ethernet, ATM, Fibre Channel, SAS, SATA,TCG SSC Encryption, IP routers, switches, etc). Familiar with protocol analyzer and traffic jammer. - Extensive hands on experience on design, development and debugging software in real-time operating systems, including Linux Kernel, Device Drivers, cooperative user space applications (IPC, socket, multi-threading etc), and substantial firmware level developments. - Solid knowledge on data structures and algorithms, computer architecture, real-time operating system, and multi-task computing. - Great analytical skills and strong problem solving ability.
Built a complete private cloud product with proprietary and open source softwares. Major open source technologies leveraged include: Linux kernel, KVM/QEMU, OpenStack, Kubernetes, Docker, Ceph, etc. The followings are some of the components I personally designed and implemented: - Distributed storage: Proprietary software (Linux LKM). Achieved 5x to 10x IO performance than open source storage - Cloud HA: Add-on features to open source softwares. Keep cloud operational in case of single point of failures in hardware or software. Including Kubernetes HA, OpenStack HA, Server networking HA, storage support for VM live migration - Cloud storage with open source storage (Ceph): Provides different storage tiers with different performance, capacity and cost profiles. Provides storage pools with different redundancy rules (replicated, erasure coding, etc). - AI/DeepStream based video analytics: Used Nvidia's DeepStream framework with GeoForce RTX2080 GPU (passthru to VM) with gstreamer, cuda 10, TensorRT, Yolo object detection algorithm, karfa, spark, cassandra, etc to extract useful metadata information from live or pre-recorded video streams
- Linux Kernel storage stack and SAS device driver hardening. - General Linux Kernel bug triage and fixing. - Routine triage of DevOps issues. - Customer escalation support.
The ultimate go-to person for any IO related technical issues encountered during product development or customer support. Hands-on projects: - Linux IO stack hardening - Data-At-Rest Encryption - Platform and disk FW online upgrade - SAS/SATA HDD/SSD/interposer error handling