Infrastructure Operations

Instructions

Operate the ITI n8n + Dify Docker stack on a daily basis. All commands assume the compose project lives at ITI/infrastructure/n8n-dify/.

Daily operations cadence:

Task	Command	Frequency
Health check	docker compose ps	Daily
Log review	`docker compose logs –since 24h \	grep -i error`	Daily
Backup	bash backup.sh	Daily
Resource check	docker stats --no-stream	Weekly
Image updates	docker compose pull && docker compose up -d	Monthly
Execution review	n8n UI > Executions > filter “failed”	Weekly
KB quality spot-check	Dify UI > Knowledge > Retrieval Testing	Weekly

Health check procedure:

1. docker compose ps — all 9 containers must show healthy or running
2. curl -s http://localhost:5678/healthz — n8n returns {"status":"ok"}
3. curl -s http://localhost:3001/console/api/setup — Dify Nginx proxy responds (status < 500)
4. redis-cli ping — returns PONG
5. docker exec iti-postgres pg_isready -U postgres — returns “accepting connections”

Backup execution and verification:

cd ITI/infrastructure/n8n-dify
bash backup.sh
ls -la /Users/username/Cursor/Archives/n8n-db-$(date +%Y%m%d).sql
ls -la /Users/username/Cursor/Archives/dify-db-$(date +%Y%m%d).sql
ls -la /Users/username/Cursor/Archives/n8n-data-$(date +%Y%m%d).tar.gz
ls -la /Users/username/Cursor/Archives/dify-data-$(date +%Y%m%d).tar.gz

Backups auto-prune after 7 days. Verify file sizes are non-zero.

Restore from backup:

docker exec -i iti-postgres psql -U postgres -d n8n < /Users/username/Cursor/Archives/n8n-db-YYYYMMDD.sql
docker exec -i iti-postgres psql -U postgres -d dify < /Users/username/Cursor/Archives/dify-db-YYYYMMDD.sql
docker run --rm -v n8n_data:/data -v /Users/username/Cursor/Archives:/backup alpine sh -c "cd /data && tar xzf /backup/n8n-data-YYYYMMDD.tar.gz"
docker compose restart

Resource monitoring:

docker stats --no-stream > /Users/username/Cursor/Archives/docker-stats-$(date +%Y%m%d).log

Thresholds: total memory < 6 GB normal, 6-8 GB acceptable, > 8 GB investigate. Largest consumers are typically iti-postgres (~1.5 GB) and iti-dify-worker (~1.3 GB).

Log rotation:

docker compose logs --since 24h > /Users/username/Cursor/Archives/n8n-dify-logs-$(date +%Y%m%d).log

n8n execution data pruning:

• Settings > Executions > set “Execution Data Prune Threshold” to 30 days
• Enable “Save Successful Executions” for auditing
• Monitor execution DB size: docker exec iti-postgres psql -U postgres -d n8n -c "SELECT pg_size_pretty(pg_database_size('n8n'))"

Dify KB re-indexing: Trigger re-indexing when retrieval quality degrades or after embedding model changes. Use Dify UI > Knowledge > select dataset > Re-index, or via API: POST /console/api/datasets/{id}/indexing-estimate.

Container restart and recovery:

• Single service: docker compose restart
• Full stack restart: docker compose down && docker compose up -d
• Nuclear reset (destroys data): docker compose down -v — only after confirmed backup
• After restart, re-run health check procedure above

Incident triage decision tree:

1. Check docker compose ps — is the container running?
2. If stopped: docker compose logs --tail 50 — check exit reason
3. If running but unhealthy: check healthcheck endpoint directly
4. If memory pressure: docker stats --no-stream — identify the consumer
5. If disk: docker system df — prune with docker system prune
6. If n8n webhooks return 500: check n8n Executions for error details
7. If Dify 502: restart dify-api and dify-nginx

Key file paths:

• Compose file: ITI/infrastructure/n8n-dify/docker-compose.yml
• Environment: ITI/infrastructure/n8n-dify/.env
• Backup script: ITI/infrastructure/n8n-dify/backup.sh
• Nginx config: ITI/infrastructure/n8n-dify/dify-nginx.conf
• Test suite: ITI/infrastructure/n8n-dify/tests/
• Archives: /Users/username/Cursor/Archives/