Scheduler Patterns

This page provides operator-friendly examples for running fyers-store under a supervisor.

Windows Task Scheduler (Example)

Use Task Scheduler to restart the process on failure.

Example command:

python src/run_store.py --auth-recovery once --poll-seconds 1

Safety notes: - Ensure FYERS_ENVIRONMENT is set correctly in the task environment. - Do not configure aggressive restart loops; allow time for token refresh.

Linux systemd (Example)

Create a service unit that restarts on failure.

Example service:

[Unit]
Description=FYERS Store Runner
After=network.target

[Service]
WorkingDirectory=/opt/fyers-store
Environment=FYERS_ENVIRONMENT=LIVE
Environment=ALLOW_LIVE_ORDERS=true
ExecStart=/usr/bin/python /opt/fyers-store/src/run_store.py --auth-recovery once
Restart=on-failure
RestartSec=15

[Install]
WantedBy=multi-user.target

Cron (Not Recommended for Live Trading)

Cron is best for batch backtests, not continuous live trading. Use a supervisor instead.

Scenario Example

Scenario: The process exits due to AuthError. What happens: - The supervisor restarts the process. What to check: - Token freshness and auth logs. What to do next: - Refresh token manually if repeated AuthError occurs.