About¶
I'm Tarak Nath Sinha, an Infrastructure & Platform Architect and SRE Engineering Leader with 16+ years of experience designing, scaling, and operating highly available infrastructure, Kubernetes platforms, distributed systems, and security-critical services. I'm a CNCF Kubestronaut holding all five Kubernetes certifications (CKA, CKAD, CKS, KCNA, and KCSA), and a Red Hat Certified Architect (RHCA).
SREOPS is where I write down what I learn from keeping real systems healthy: how clusters behave under pressure, why alerts fire, and how to recover quickly when something breaks. Everything here comes from hands-on work, not slideware.
What I do¶
- Kubernetes operations: building, upgrading and securing clusters that stay boring in the best way.
- Reliability engineering: defining SLOs, and building monitoring and alerting that wakes people only when it matters.
- Incident response and recovery: runbooks, backups and restore drills that have been tested before the outage, not during it.
- Linux and automation: repeatable, version-controlled infrastructure instead of one-off fixes.
Background¶
With 16 years of experience in infrastructure engineering, Site Reliability Engineering, and platform operations, I've worked across enterprise technology environments, cloud-native platforms, and large-scale production systems. My experience spans Kubernetes, Linux, automation, networking, observability, disaster recovery, and security-focused infrastructure.
I currently lead SRE engineering teams, focusing on platform reliability, operational excellence, and engineering practices for critical production services.
Why Kubestronaut¶
Kubestronaut is CNCF's recognition for people who have passed every Kubernetes certification it offers. Three of those exams are hands-on, performance-based tests in a live terminal, so the title reflects practical skill, not memorised answers. See the full credential breakdown.
How I work¶
- Measure before changing. Decisions come from metrics and logs, not guesses.
- Automate the second time. The first fix is manual; the repeat becomes code.
- Write it down. A runbook that exists beats expertise that lives in one person's head.
- Design for failure. Everything breaks eventually, so I plan how it breaks.
Let's talk¶
If you want a second pair of eyes on your cluster, your alerting, or your recovery plan, see how I can help or connect with me on LinkedIn.