NinjaTrader after a Windows update: verify startup and strategy recovery

Identify what actually returned after the reboot

Inspect the machine before manually opening apps or enabling strategies. Starting them yourself can make the desktop look recovered while hiding the startup step that failed.

Check Windows sign-in and the two startup settings

Windows startup applications normally run when a user signs in. A machine waiting at the sign-in screen has not reached that stage. If applications appear only after you log in remotely, record that dependency; remote access succeeding is not evidence that recovery started unattended.

In Ninja Watchdog, inspect both "Start Ninja Watchdog with Windows" and "Launch NinjaTrader after Windows starts." The first starts the monitoring app after the Windows account signs in; the second asks it to open NinjaTrader. Windows sign-in and NinjaTrader sign-in are separate steps.

If Ninja Watchdog is missing after sign-in, inspect its configured startup method. The current application uses a logon-triggered Task Scheduler entry, with a current-user startup entry as a fallback. Check the task as well as Windows Settings > Apps > Startup; absence from one list does not establish that startup is unconfigured. Confirm the installed executable path and the Windows account used for trading.

If Watchdog starts but NinjaTrader stops at a dialog, capture the actual prompt. A login request, component error or NinjaWatchdogBridge authorization request requires a different response from a slow launch. The bridge must be allowed for recovery to operate.

Match the strategy to the account that should be running

Open the Control Center Strategies tab. If the row is missing, review its active-only and account filters before adding a second instance. Match the account name, instrument and expiry, Data Series, workspace and parameters against the intended setup. A descriptive account label can differ from the actual account name.

Read the row state as well as its presence: a black strategy name is disabled, orange is waiting for the strategy position to become flat, and green is running. A waiting strategy is different from an app that failed to launch. Use the dedicated restart guide for persistence and state diagnosis.

For multiple connections, inspect the provider used by that account in the Connections menu. The lower-left status is aggregated; a restored connection elsewhere does not explain the selected account. Check price and order connectivity separately when the platform reports a loss.

Before re-enabling, review existing account positions and working orders alongside the strategy position and Start Behavior. Synchronization options can place live orders. Do not switch to an immediate-start mode merely to remove a waiting state.

Verify the unattended path with separate timestamps

For a planned restart in simulation, note when the desktop becomes available, Ninja Watchdog appears, NinjaTrader connects and the intended strategy reaches its expected state. Observe the configured startup path before intervening. This reveals whether time was spent at sign-in, application loading, connection recovery or strategy startup.

Compare the recovery email with the same event and machine, then check the current strategy state. Email arrival establishes delivery of that report; the strategy grid establishes what is running now. An old report or a successful application launch cannot stand in for that check.

Repeat this observation after an actual Windows update if it changes prompts or startup behavior. A normal reboot test does not exercise every update-specific condition.

Set the update window around the automation schedule

On Windows 11, open Settings > Windows Update > Advanced options > Active hours. Choose the period when the machine needs to remain available, including unattended strategy hours. An automatically inferred period may not match your trading schedule.

Active hours help reduce inconvenient update restarts; they do not verify application recovery. For a managed VPS, also check the provider's maintenance schedule and leave time afterward for the observed startup path. Avoid making the first post-update launch coincide with the moment a strategy needs to run.