• A systems-oriented DevOps Engineer who goes beyond tool-level usage to dissect the internals:

  • Kubernetes networking at the Linux kernel level (iptables, route tables, conntrack, Calico CNI)
  • Nginx event-driven architecture, and kernel system calls to solve problems at their root.
  • Prometheus TSDB internals, optimizing kernel writeback and WAL/mmap to mitigate I/O spikes and ensure high-throughput zero-copy storage.

• Actively documenting the journey on a tech blog with 300+ posts and 130K+ visitors.

• Selected (Nov 2025) as the founding AWS Cloud Club Captain for Konkuk University.

✉️ Email

mr8356@naver.com

🐱 Github

https://github.com/mr8356

📄 Blog

동현의 코딩여행 (konkukcodekat.tistory.com)

LinkedIn

http://www.linkedin.com/in/mr8356

• Language: Java(Spring Boot), Python(PS), C(Device Driver)
• Cloud-Native: AWS (Lambda, EKS, DynamoDB, RDS, EC2, Api-gateway, ELB, SQS), AWS CLI
• Infrastructure: Kubernetes, Docker, OpenStack, Nginx, Terraform(Basic)
• Monitoring: Prometheus, Grafana, Istio
• CI/CD: GitHub Actions
• DB: MySQL

• Konkuk University | Computer Science & Engineering
• Mar 2021 - Feb 2027 (Expected)
• GPA: 4.26 / 4.5
• Academic Excellence Scholarship (Top 5%)
  • 2024 Fall Dean's List
  • 2025 Spring Dean's List
  • 2025 Fall President's List

• [K8s] Kernel-Level Networking Analysis: Service VIP to Pod (iptables, route, Calico CNI IPIP Mode)
• [DevOps Interview Retrospective] TLB, K8s Internals & Kernel Isolation, TLS Handshake
• [DevOps Interview Retrospective #2] K8s Control Plane Scalability & Load, mTLS
• Limitations of Thread-Based Servers (C10K Problem) & Nginx Process Model for Reducing Context-Switching Overhead
• How Nginx Handles High-Volume Traffic via Non-blocking I/O Multiplexing & epoll System Calls
• [Nginx] Zero Copy (sendfile) Kernel Internals & System Call Tracing via strace

"Zero-Downtime Migration & 87% Infrastructure Cost Reduction Without a Dedicated Server Admin"

Period: Mar 2026 (Short-term Contract)

• Role: FinOps Consultant — Infrastructure Diagnosis, Zero-Downtime Migration, Architecture Redesign
• Tech: AWS EC2, EBS, CloudWatch, EC2 Image Builder, ALB, VPC, S3, CloudFront, Route 53, ACM, IAM
• Blog: https://konkukcodekat.tistory.com/377

1. FinOps in Practice: 83% EC2 Cost Reduction ($84 → $14)
   • [Analysis] Proved overprovisioning of t2.large with CloudWatch CPU metrics (avg 1%, peak <20%).
   • [Decision] Persuaded the CEO to adopt t2→t3 migration + 3-year RI prepayment (ROI 165%), reducing monthly cost from $84.10 → $14.30 (83% savings).
2. Deep Dive: Zero-Downtime EC2 Spec Downgrade
   • [Problem] Server admin had resigned; stopping the instance risked unrecoverable service outage.
   • [Solution] EBS Snapshot → EC2 Image Builder AMI Pipeline → ALB switching achieved 0 minutes of downtime. Worked around IAM Trust Policy omission and vmie kernel version mismatch (6.8.0-1040-aws) bug via native AMI manual creation.
3. Over-Engineering Discovery & Serverless Migration
   • [Analysis] Discovered 7 zombie EIPs via VPC ENI/ELB dependency analysis. Confirmed static HTML was being served via EC2+ALB.
   • [Result] Migrated to S3 + CloudFront CDN + ACM SSL + Route 53 Alias (Serverless). Total infra cost: $115+/mo → $14/mo (87% reduction), saving $3,600+ over 3 years.

"70% Cost Reduction & Data Pipeline Optimization via Serverless Architecture"

Period: Sep 2025 - Present

• Role: Cloud Architecture Design & Backend Development, IaC Implementation (In Progress)
• Tech: AWS Lambda, Athena, Glue, S3, EventBridge, Bedrock, SQS, Terraform
• GitHub: https://github.com/FoodDonorService/foodDonor-aws

1. FinOps in Practice: Data-Driven Serverless Migration (70% Cost Reduction)
   • [Analysis] EC2 deemed inefficient for intermittent ETL workloads; performed cost simulation with AWS Calculator.
   • [Decision] Quantitatively proved Serverless (Lambda, Athena) is 70% cheaper than EC2, and built an MSA architecture based on API Gateway + Lambda for both cost efficiency and flexibility.
2. Deep Dive: Ensuring Stability for Large-Scale Processing
   • [Problem] Lambda Runtime OOM (Out Of Memory) during large-scale JSON ingestion.
   • [Solution] Instead of costly resource scaling, applied Streaming & Chunking techniques to stabilize the pipeline through logic optimization alone.
     • [AWS Lambda] Streaming + Chunking to Resolve Runtime.OutOfMemory in Big Data Batch Pipeline
3. Athena Query Performance & Cost Optimization (90% Cost Reduction)
   • [Optimization] Identified cost leakage by analyzing Athena's billing model (scanned data volume).
   • [Result] Converted data to Parquet via Glue and applied S3 Partitioning, blocking unnecessary scans to achieve 90% query cost reduction.
     • [AWS Glue] PySpark: Converting Streamed Split JSON to Columnar Data via Parquet + Partitioning
4. Legacy Infrastructure → Terraform IaC Migration (Reverse Engineering)
   • [Modernization] Reverse-engineered console-based infrastructure using Former2, removed hardcoding, modularized, and set up S3 Backend.

"2.5x Performance Improvement via Kubernetes-Based Auto-Scaling"

Period: Dec 2024 - Jun 2025

• Role: Infrastructure Lead & Backend Engineer
• Tech: Kubernetes, Java, Spring Boot, Redis
• GitHub: https://github.com/mr8356/cash_ticket

1. Kubernetes Infrastructure Setup & Operations (Ownership)
   • [Challenge] Single-server architecture could not guarantee service stability during traffic spikes. Despite the steep learning curve, personally built and operated a Kubernetes cluster to establish a high-availability (HA) architecture.
   • [Result] Applied HPA (Horizontal Pod Autoscaler) to build a system that flexibly responds to traffic fluctuations.
2. Redis Queue System for Traffic Control
   • [Optimization] Used Redis Sorted Set to prevent DB connection exhaustion during first-come-first-served events, maintaining O(log N) lookup complexity for high-performance queue design.
   • [Performance] Prevented system downtime via flow control; Locust load testing showed 2.5x improvement in throughput over baseline.
3. Data Consistency & Code Quality Assurance
   • [Reliability] Introduced Spring Scheduler (CronJob) for periodic cleanup of expired tickets to ensure data consistency. Led regular Code Reviews to minimize technical debt.

"Minimized System Coupling via Event-Driven Architecture"

Period: Dec 2024 - Jun 2025

• Role: Server Lead & Architect
• Tech: Spring Boot, Redis MQ, Docker, GitHub Actions
• GitHub: Roomie-Server (github.com/WeAreRoommies/Roomie-Server)

1. Bottleneck Resolution via EDA Adoption
   • [Issue] Synchronous move-in application process caused cascading failures and latency spikes under high traffic.
   • [Architecting] Defined business requirement: "System stability takes priority over notification immediacy." Introduced Redis MQ-based EDA.
   • [Result] Decoupled core logic from auxiliary logic asynchronously, reducing coupling and achieving 1.5x improvement in write throughput (nGrinder benchmark).
2. CI/CD Pipeline Implementation
   • [Automation] Aimed to eliminate operational inefficiency from repetitive manual deployments.
   • [Impact] Fully automated the deployment pipeline using Docker and GitHub Actions, reducing deployment time and preventing human errors.

"Built Private Cloud on OpenStack & Established Observability Environment"

• Lab Page: DongHyun Jo (dms.konkuk.ac.kr)
• Blog: K8s Monitoring: Troubleshooting NetworkPolicy Firewall & ServiceMonitor Issues with Separate Helm Charts for Loki and Prometheus

Period: Dec 2024 - Feb 2025

1. IaaS Environment Setup & Deep Dive
   • [Action] To deeply understand cloud virtualization principles, built a private cloud from bare-metal using OpenStack.
   • [Outcome] Organized and led an OpenStack study group; completed IaaS environment setup and documented the entire process on technical blog (OpenStack & Kubernetes categories).
2. Observability
   • [Problem] No observability tooling to understand the state of the deployed infrastructure.
   • [Solution] Integrated Prometheus and Grafana to build a real-time monitoring dashboard for the hybrid cloud environment.

2025.11 - Present

• Selected as the inaugural AWS Cloud Club Captain at Konkuk University (Founding Captain).

2025.09 - Present

• Designed and delivered a practice-oriented intensive 3-week server curriculum, contributing to engineer training.
• https://youtube.com/playlist?list=PL7z9i3yYqDwyQDKdaqjTQoFfEuPWlnztL&si=t4y4IzHrZTIUW0Cn

2025.01 - 2025.12

• Led 15-week, 30-hour weekly seminars covering Git, SW Architecture, Cloud Systems, Microservices Architecture, Monitoring, and SA.
• https://konkukcodekat.tistory.com/category/My seminar/Bibimbap(President)

2024.09 - 2025.02

• Immediately after joining as a sophomore, selected as server lead for 3 major events: Joint Seminar, Hackathon, and Demo Day.
• Led server development for the 'Roomie' project.

• SOPT 35th Hackathon — Grand Prize
• SOPT 35th Demo Day — Excellence Award
• 2025 Summer CS Department Hackathon — Excellence Award