[ ok ] mounting /dev/core ...

[ ok ] starting kubelet (412d uptime)

[ ok ] verifying tls handshake

[ warn ] dazzle.fx loading ...

[ ok ] welcome, operator.

SYS://YRSHENG.CORE · v2026.06 · STATUS: ONLINE

yrsheng.

Senior DevOps Engineer.
Orchestrating scale. Automating perfection at 3 a.m. so you don't have to.

hire me → view toolkit ↓
Cluster Uptime
0%
Deployments / yr
0
Nodes Orchestrated
0
Mean Recovery
0s
// 01 · TOOLKIT

Mastery over the
modern infrastructure stack.

Battle-tested. Production-grade. Built to never wake you up.

Kubernetes Native

Multi-cluster federation, custom operators, GitOps via ArgoCD/Flux, self-healing workloads at 400+ node scale.

k8s · helm · argocd

Linux Deep Systems

Kernel tuning, eBPF observability, perf profiling, custom systemd units. From bootloader to user-space.

kernel · bpf · perf

Aliyun Architect

VPC topology, ECS auto-scaling, ACK, OSS pipelines, RAM hardening — resilient cloud at predictable cost.

vpc · ack · oss · slb

Python Automation

FastAPI services, asyncio pipelines, monitoring agents, infra-as-code generators. Glue that never breaks.

fastapi · asyncio · click

CI/CD & GitOps

Jenkins, GitLab CI, GitHub Actions. Commit→prod in < 4 min. Progressive delivery, zero-downtime, no drama.

jenkins · gitlab · actions

Observability

Prometheus, Grafana, Loki, OpenTelemetry. SLOs that make sense; alerts that actually mean something.

prom · loki · otel
// 02 · SERVICES

Engineering services
that pay rent.

Transforming infrastructure into a measurable advantage — not a cost center.

/ 001

Cluster Reliability

I build and run enterprise-grade Kubernetes clusters with self-healing, autoscaling, and DR drills. Your platform stays up — even on the bad days.

Discuss clustering →
/ 002

CI/CD Automation

Commit-to-prod in minutes, not days. Bulletproof release pipelines with progressive rollout, automated rollback, and audit-ready trails.

Streamline delivery →
/ 003

Cloud Cost Optimization

Right-size workloads, kill zombie resources, build chargeback dashboards. Same SLOs, 30–50% lower spend — without anyone noticing.

Cut the bill →
/ 004

Observability & SRE

From RED metrics to error budgets. SLO-driven engineering so the team ships fast — and pages stay quiet.

Wire up signals →
// production · live tail streaming
    // 03 · CONTACT

    Let's talk scale.

    Available for contract roles, platform consulting, and rescue missions. Reach out — I read everything.

    📍 Shanghai · CN
    ✉ [email protected]
    ⏱ avg. reply < 4h