Docker Config QA Regression

Docker Config Change QA & Regression Pattern

Context: After modifying Docker container config, volume mounts, or environment variables — systematic verification to catch cascading failures

Problem

Docker config changes (volume mounts, env vars, compose edits) can cause cascading failures that aren't immediately obvious. A config fix can break auth, lose credentials, or disconnect services. Without systematic QA, you discover these in production.

Solution — QA Regression Checklist

Run this systematic check after ANY docker-compose.yml or config change:

Phase 1: Container Health

# Container running?
docker ps --format 'table {{.Names}}\t{{.Status}}\t{{.Ports}}' | grep <name>

# Clean startup? (no errors)
docker logs --tail 30 <container> 2>&1

# Errors since latest start only (filter by timestamp)
START_TIME=$(docker inspect <container> --format '{{.State.StartedAt}}' | cut -c1-19)
docker logs <container> 2>&1 | awk -v ts="$START_TIME" '$0 >= ts' | grep -iE 'error|fail|crash' || echo 'CLEAN'

Phase 2: Volume Mounts

# All mounts present?
docker inspect <container> --format '{{range .Mounts}}{{.Source}} -> {{.Destination}} ({{.Type}})
{{end}}'

# Host and container in sync?
diff <(md5sum /host/path/config.json) <(docker exec <container> md5sum /container/path/config.json)

Phase 3: Credentials & Auth

# Auth profiles exist?
docker exec <container> find / -name 'auth-profiles*' -o -name 'credentials*' 2>/dev/null

# API keys loaded?
docker exec <container> env | grep -E 'API|KEY|TOKEN|SECRET' | sort

Phase 4: Config Integrity

# Critical config values unchanged?
docker exec <container> cat /path/config.json | jq '{
  key_setting_1: .path.to.setting1,
  key_setting_2: .path.to.setting2
}'

Phase 5: Service Connectivity

# Dashboard/web UI accessible?
curl -s -o /dev/null -w 'HTTP %{http_code}' http://localhost:<port>/

# External services connected? (check logs for provider startup)
docker logs <container> 2>&1 | grep -i 'starting provider\|connected\|ready'

Phase 6: Persistence Test

# Save current state
docker exec <container> cat /path/config.json | jq '.critical.field' > /tmp/before.txt

# Full destroy/recreate cycle
docker compose down && docker compose up -d
# Wait for startup...

# Compare
docker exec <container> cat /path/config.json | jq '.critical.field' > /tmp/after.txt
diff /tmp/before.txt /tmp/after.txt && echo "PERSISTED" || echo "REGRESSION!"

Key Lesson: Auth Files Are Often Missed

When adding new volume mounts, the most commonly missed files are:

• Auth profiles / API key stores (nested in agent/service subdirectories)
• Session tokens / offset files (e.g., Telegram update offsets)
• Cron job state (scheduled task tracking)

Always run find on the old config path and compare with the new one.

When to Use

• After modifying docker-compose.yml (volumes, ports, env)
• After editing config files that affect container behavior
• After any docker compose down && up cycle
• Before declaring a Docker infrastructure change "done"