Troubleshooting

Resolution guide for common issues. If you don't find your situation here, contact hello@myeline.io and attach:

• Output of curl -s https://your-domain/health
• Last 100 lines of the web container logs (podman logs --tail 100 myeline-web)
• Installed version: curl -s https://your-domain/healthz | jq .version

Sections

• Licence errors — LicenseRevokedError, LicenseExpiredError, invalid signature, clock skew.
• Ollama issues — model not loaded, OOM, excess latency, GPU not detected.
• Sync failures — OAuth refused, expired token, provider 429, ignored files.

Quick diagnostic

# 1. Global state
curl -s https://your-domain/health | jq

# 2. Recent logs
podman logs --tail 50 myeline-web
podman logs --tail 50 myeline-cron
podman logs --tail 50 myeline-worker

# 3. Dependency connectivity
podman exec myeline-web flask diagnose

# 4. Detected deployment mode
podman exec myeline-web python -m app.utils.deployment_mode

flask diagnose (internal command) attempts a connection to DB / Redis / Ollama / mailer / AI provider and prints a synthesised report.

When to restart?

Symptom	Action
Web no longer responds, healthz timeout	podman restart myeline-web
RQ workers stuck, queue grows	podman restart myeline-worker
Cron skipped without reason	podman restart myeline-cron
Slow embedding + GPU full	podman restart ollama
.env change	podman-compose down && podman-compose up -d (cf. CLAUDE.md)
DB migration after git pull	podman exec myeline-web flask db upgrade

A simple podman-compose restart web does not reload environment variables — always down + up after a .env change.