March 2026: Wartime Challenges & Staging Environment Planning

March 2026 presented unique challenges - building Goalixa during wartime in Iran with severe internet restrictions. Despite connectivity issues, significant progress was made in infrastructure planning and development workflow improvements.

Major Theme: Building Under Crisis

🇮🇷 Wartime Internet Restrictions

The most significant challenge in March was the internet situation in Iran. Most people had no stable connection, and only a small number could connect using expensive VPN services.

Personal Posts:

Survive in War Time (March 6)
Learning in Wartime: Preparing Infrastructure (March 11)

Challenges Faced:

Intermittent internet access
Cannot access cluster when internet is down
Cannot deploy changes or fix production bugs quickly
Simple engineering tasks become difficult

Adaptation Strategy:

Use downtime for deep learning and study
Plan infrastructure improvements thoroughly
Focus on what can be done offline
Prepare for when connectivity returns

Learning Topics (during downtime):

Kubernetes deeper internals
Cilium for CNI
Envoy and service mesh patterns
Production-grade cluster operations

Infrastructure Improvements

🔧 Staging Environment Plan (March 11)

With every code change going straight to production and causing outages, planning began for a staging environment.

Key Post: Why Goalixa Needs Staging

Current Problems:

No staging on Kubernetes cluster
Every deploy risks production
Buggy changes take the site down
Hotfixes happen directly on prod (risky)

Staging Goals:

Production-like environment for pre-production testing
Safe environment to break and learn
Predictable promotion: dev → staging → production

Approach: Kustomize Overlays

Base: Shared manifests for all components
Overlays: staging and production for environment-specific differences
Image Tags: Same build, promoted by tag not rebuild
Config Parity: Keep env vars/ConfigMaps aligned
Ingress: Separate hostnames (staging.goalixa.com)
Secrets: Same secret manager path pattern

Parity Checklist:

Same services (PWA, API Gateway/BFF, Core API, Auth)
Same resource types and naming
Same ingress class and TLS approach
Environment-specific values in overlays

Development Workflow Improvements

🤖 AI-Assisted Development (March 12)

Implemented better workflows for using Claude effectively to speed up Goalixa development.

Key Post: Using Claude to Speed Up Development

Problems Solved:

Every Claude session needed long context
Auto-generated CLAUDE.md files were low quality
AI-written code may need refactor but needed quick fixes

Solution: Service-Level CLAUDE.md Each microservice now has a CLAUDE.md with:

What the service does and main endpoints
Stack, build, test, lint/format commands
Run/dev steps and sample .env
Deploy info (image, tag, kustomize/helm path)
Contracts (DB/queues/external APIs)
High-risk areas and tech debt
Sample prompts for common tasks

Standard Session Prompt: “Use CLAUDE.md, follow coding standards, don’t touch secrets, run tests with npm test”

Rules for AI Changes:

Keep changes small: one feature/fix per request
Demand test plan and self-review checklist
Review all AI-generated code before committing

February Carryover: Latency Lessons

While the latency incident happened in late February, March brought deeper reflection on the engineering lessons.

Key Post: Latency Taught Me Better Software Engineering

Engineering Mindset Shift:

Design for the full request path (not just one service)
Keep contracts simple and stable
Add observability before scaling complexity
Fix bottlenecks before adding more components
Cache intentionally where safe and valuable

Current Focus:

Stabilize PWA, API Gateway, and Core API
Remove avoidable latency from routes
Enable caching after bug fixes stabilize

Infrastructure Plans (When Connectivity Returns)

📋 Planned Upgrades

When stable internet access returns, the focus will be on:

Kubernetes Cluster Setup
- Stand up full cluster
- Add Cilium for networking and security
- Implement service mesh with Envoy
- Adopt professional-grade tooling
Staging Environment
- Implement Kustomize overlays
- Create staging.goalixa.com
- Establish promotion pipeline
- Reduce production risk
Monitoring Expansion
- Add more per-service metrics
- Implement SLI/SLO-based alerts
- Create dashboard views for incident triage
- Define severity levels and escalation

Statistics

Posts Published: 4 posts Major Challenges: Internet restrictions, wartime conditions Infrastructure Planning: Staging environment designed Development Workflow: AI-assisted development improved Learning Focus: Kubernetes, Cilium, service mesh

Key Takeaways

Resilience: Building under crisis requires adaptability and planning
Downtime Productivity: Use connectivity interruptions for learning and planning
Staging Necessity: No staging environment = production incidents
AI Workflow: Structured approach to AI-assisted development improves quality
Infrastructure Debt: February’s migration success means March focuses on reliability

Personal Reflection

March was emotionally and technically challenging. Building a startup is hard enough without wartime conditions. However, the situation reinforced the importance of:

Resilient infrastructure that can withstand connectivity issues
Staging environments to prevent production incidents
Observability to debug issues when they occur
Planning for when connectivity returns

The learning during downtime (Kubernetes, Cilium, service mesh) will be valuable when implementing these improvements.

Looking Ahead to April

April Goals:

Implement staging environment with Kustomize
Deploy Cilium for cluster networking
Begin service mesh exploration with Envoy
Expand monitoring and alerting coverage
Continue building (as internet access allows)

Priority: Stabilize infrastructure and reduce deployment risk before adding new features.

Previous Month: See February 2026 Recap for the major migration milestone.

February 2026 API Gateway Pwa Microservices