# 11 - RouteTrack Pi — Logger Service Cleanup & Boot Reliability

#### **Date:** December 25, 2025  
**Category:** Raspberry Pi / systemd / GPS / Reliability  
**Backlink:** [10 – RouteTrack Pi — Shift Mode (SQLite + Flask API + Dashboard Controls)](https://docs.natenetworks.com/books/06-raspberry-pi-python-linux-tips/page/10-routetrack-pi-shift-mode-sqlite-flask-api-dashboard-controls)

---

## Project Goal

Before taking RouteTrack mobile (car/truck use), I wanted to ensure the system behaves like an appliance:

- Power on → services come up automatically
- No manual “status checks” required every boot
- Logging survives reboots and unexpected shutdowns
- Logging is independent of network availability (portable use)

This entry documents the final cleanup to the `routetrack-logger.service` unit so it is:

- Dependency-safe (waits for GPSD)
- Restart-safe (always recovers)
- Logging-clean (journald only, no conflicting directives)

---

## Problem Identified

The logger service originally contained **conflicting output directives**, which can create confusion about where logs are actually going.

Example conflict pattern:

- `StandardOutput=append:/path/to/file.log`
- followed later by `StandardOutput=journal`

In systemd, the **last directive wins**, meaning file append logging may silently stop even though it appears configured.

To keep RouteTrack stable and simple, the logger service was standardized to **journald-only** logging.

---

## Final Logger Service Configuration (Clean + Portable)

### View the service

```bash
sudo systemctl cat routetrack-logger.service

```

### Final contents used

```ini
# /etc/systemd/system/routetrack-logger.service

[Unit]
Description=RouteTrack GPS Logger

# Do not start logger until gpsd is up
After=gpsd-standalone.service network.target
Wants=gpsd-standalone.service
Requires=gpsd-standalone.service

[Service]
Type=simple

# Run from project root
WorkingDirectory=/opt/routetrack

# Use the Python venv so RouteTrack is isolated from system packages
ExecStart=/opt/routetrack/venv/bin/python /opt/routetrack/bin/routetrack-logger.py

# Restart forever on crash/disconnect
Restart=always
RestartSec=3

# Send stdout/stderr into journald
StandardOutput=journal
StandardError=journal

[Install]
WantedBy=multi-user.target

```

---

## Why This Unit File Is “Truck Safe”

### ✅ GPS Dependency Enforcement

Because of:

```ini
Requires=gpsd-standalone.service
After=gpsd-standalone.service

```

systemd will not attempt to start RouteTrack logging until GPSD is up.

---

### ✅ Crash Recovery

Because of:

```ini
Restart=always
RestartSec=3

```

if the logger crashes or the GPS temporarily drops, RouteTrack automatically restarts.

---

### ✅ Logging That Doesn’t Break

Because of:

```ini
StandardOutput=journal
StandardError=journal

```

all logs are always available via systemd journald, which is reliable even with power cycling.

---

## Reload & Restart Procedure

After any systemd file changes:

```bash
sudo systemctl daemon-reload
sudo systemctl restart routetrack-logger.service

```

---

## Confirm Boot Enablement

The portable goal is to skip status checks entirely, so the services must auto-start on boot.

Verify enablement:

```bash
systemctl is-enabled gpsd-standalone.service routetrack-logger.service routetrack-dashboard.service

```

If any return `disabled`, enable them:

```bash
sudo systemctl enable gpsd-standalone.service routetrack-logger.service routetrack-dashboard.service

```

---

## “No Status Checks” Workflow

Once enabled:

- If the dashboard loads from my phone → dashboard service is running
- If I can press **Start Shift** → API is live
- If route points update while driving → GPSD + logger are working

This allows RouteTrack to behave like a real vehicle appliance system.

---

## Quick Troubleshooting (Single Command)

If anything seems off after boot, the first and best check is:

```bash
journalctl -u routetrack-logger -n 50 --no-pager

```

---

## Result

RouteTrack is now ready for mobile testing:

GPS logger waits for GPSD  
Logger restarts automatically if anything fails  
Journald logging is clean and consistent  
Boot enablement supports “power on and go” use

---

## Next Steps

The next phase is real-world validation:

- Start a shift before leaving
- Drive a short route (gas station test)
- End shift on return
- Run processor and confirm stop events + summaries update