Xây dựng hệ thống quan sát, tự động hóa nền tảng và tối ưu hiệu năng hệ thống chịu tải cao cho Zalo.
Yêu cầu
3+ năm kinh nghiệm kỹ thuật phần mềm tầng hệ thống với C++, Go, Java hoặc Rust
Hiểu sâu về hệ thống phân tán (Raft, Paxos, sharding, CAP, tính nhất quán)
Hiểu sâu về Linux Kernel (Virtual Memory, Page Cache, I/O schedulers, Syscalls)
Thành thạo giao thức mạng TCP/IP, socket, HTTP/gRPC và tối ưu độ trễ
Kiến thức sâu về Observability (Prometheus, Grafana, OpenTelemetry, Jaeger)
Trách nhiệm
Thiết kế, phát triển hệ thống Observability và bộ công cụ tự động hóa
Nghiên cứu tối ưu hóa hiệu năng hệ thống cấp thấp như Disk I/O, Linux Kernel
Chịu trách nhiệm về SLA, độ sẵn sàng, độ trễ và khả năng mở rộng dịch vụ
Chỉ huy ứng phó sự cố production, phân tích nguyên nhân gốc rễ và postmortem
Có thì tốt
Tư duy thuật toán và cấu trúc dữ liệu giải quyết bài toán phân tán phức tạp
Thành thạo công cụ chẩn đoán kernel như eBPF, perf, bcc-tools, sysctl
Kinh nghiệm thực tế với Docker, viết Custom Kubernetes Operator và IaC
Tóm tắt bằng AI từ JD gốc, đọc JD gốc trước khi ứng tuyển.
Internal Platform & Systems Engineering:
Design and develop deep Observability systems (high-cardinality metrics, distributed tracing, large-scale logging) and Automation toolkits (Control Planes, Custom Kubernetes Operators, IaC) to drive Toil Elimination;
Research, design, and optimize low-level system performance (Disk I/O, Network Stack, Linux Kernel Tuning) for large-scale distributed services.
Production Operations & Incident Response:
Take ownership of the Availability, Latency, and Scalability of production services using SLOs/SLIs and Error Budgets;
Foster a resilience for failure mindset to lead incident response for complex production issues, conduct Root Cause Analysis (RCA), and facilitate blameless postmortems to prevent recurrence.
3+ years of Software Engineering experience working directly with low-level system components such as Disk I/O, Network Stack, Memory Management, etc., using systems programming languages (C++, Go, Java, or Rust);
Distributed Systems: Strong understanding of consensus algorithms (Raft, Paxos), replication topologies, data partitioning (sharding / consistent hashing), CAP theorem, and consistency models (eventual vs. strong consistency);
Linux Kernel: Deep understanding of Linux kernel internals (Virtual Memory, Page Cache, I/O schedulers, System calls, Process/Thread management);
Networking: Proficient in TCP/IP stack, socket programming, HTTP/gRPC, load balancing, and network latency optimization at the Transport/Application layers;
Observability: In-depth knowledge of observability architecture and tools (Prometheus, Grafana, OpenTelemetry, Jaeger), handling high-cardinality metrics and large-scale logging/tracing;
SRE Mindset: "Operations as Code" mindset, reliability management via metrics (SLO/SLI, Error Budget), toil elimination, and automated safety guardrails (Progressive Rollout, Zero Touch Prod).
Nice to have:
Strong Algorithmic & Data Structure Thinking: Proven ability to solve complex data structure and algorithmic optimization problems in distributed systems;
Profiling & Diagnostic Tools: Proficiency with kernel-level diagnostic and profiling tools such as eBPF, perf, bcc-tools, and sysctl;
Container & Orchestration: Hands-on experience with Docker, Kubernetes (writing Custom Kubernetes Operators), and Infrastructure-as-Code (Terraform/Ansible).
At Zalo, we define “SRE as what happens when a Software Engineer is tasked with solving operational and infrastructure challenges.”
The Site Reliability Engineer (SRE) role requires a blend of low-level system engineering skills and a solid SRE mindset. You will leverage software engineering principles and code to build automation platforms, optimize disk/network I/O, and maintain high-scale distributed systems with high availability and ultra-low latency at millions of QPS.
At Zalo, we define “SRE as what happens when a Software Engineer is tasked with solving operational and infrastructure challenges.” The Site Reliability Engineer (SRE) role requires a blend of low-level system enginee...