FiveM Server Keeps Crashing: How to Keep It Online 24/7 with Auto-Restart

Check if it was an OOM kill (out of memory):

sudo dmesg -T | grep -i "killed process"
sudo journalctl -k --since "24 hours ago" | grep -i oom

If you see a line like Killed process 12345 (FXServer) total-vm:8GB, the kernel killed it due to lack of RAM. This points to a memory leak in a resource or an undersized VPS.

Check if it was an internal FXServer crash:

tail -n 500 /home/fivem/server/server.log | grep -iE "error|exception|crash"

Look for a stack trace or lines like [script:resource_name] ERROR. Lua resources with infinite loops or unhandled exceptions take down the whole server.

Check network saturation (possible DDoS):

sudo apt install -y vnstat
vnstat -l -i eth0

If inbound traffic (rx) is in the hundreds of Mbps with few players connected, it’s a flood. Note the peak so you can compare later.

Create a dedicated user for the server (if it still runs as root, this is mandatory for security):

sudo useradd -m -s /bin/bash fivem
sudo chown -R fivem:fivem /home/fivem/server

Running internet-facing services as root is a full-compromise vector if an RCE is ever discovered in FXServer.

Create the service unit file:

sudo nano /etc/systemd/system/fivem.service

Paste the content below, adjusting the paths to match your installation:

[Unit]
Description=FiveM FXServer
After=network-online.target
Wants=network-online.target

[Service]
Type=simple
User=fivem
Group=fivem
WorkingDirectory=/home/fivem/server
ExecStart=/home/fivem/server/run.sh +exec server.cfg
Restart=always
RestartSec=10
StartLimitBurst=5
StartLimitIntervalSec=300
StandardOutput=append:/home/fivem/server/server.log
StandardError=append:/home/fivem/server/server.error.log
LimitNOFILE=65535

[Install]
WantedBy=multi-user.target

The critical directives: Restart=always restarts on any exit (crash or normal), RestartSec=10 waits 10 seconds between attempts (gives the UDP socket time to release), and StartLimitBurst=5 prevents an infinite loop if the server is truly broken.

Enable and start the service:

sudo systemctl daemon-reload
sudo systemctl enable fivem.service
sudo systemctl start fivem.service
sudo systemctl status fivem.service

The output should show active (running) in green. If it shows failed, run sudo journalctl -u fivem.service -n 50 to see the error.

Create a systemd timer for a daily restart at 6 AM:

sudo nano /etc/systemd/system/fivem-restart.timer

[Unit]
Description=Restart FiveM daily at 6am

[Timer]
OnCalendar=*-*-* 06:00:00
Persistent=true

[Install]
WantedBy=timers.target

And the service that the timer triggers:

sudo nano /etc/systemd/system/fivem-restart.service

[Unit]
Description=Restart FiveM service

[Service]
Type=oneshot
ExecStart=/bin/systemctl restart fivem.service

Enable the timer:

sudo systemctl daemon-reload
sudo systemctl enable fivem-restart.timer
sudo systemctl start fivem-restart.timer
sudo systemctl list-timers fivem-restart.timer

The output should show the next scheduled run. You can adjust the time to another window — always pick when your server has the fewest players online.

Install htop and vnstat for real-time observation:

sudo apt install -y htop vnstat
sudo systemctl enable --now vnstat

htop shows process by process. Filter with F4 and type FXServer to see only what matters.

Create a simple alert script that checks FXServer RAM every 5 minutes:

sudo nano /usr/local/bin/fivem-mem-check.sh

#!/bin/bash
THRESHOLD=85
MEM_PERCENT=$(ps -o %mem= -C FXServer | head -n1 | tr -d ' ')
MEM_INT=${MEM_PERCENT%.*}

if [ "$MEM_INT" -gt "$THRESHOLD" ]; then
  echo "$(date): FXServer using ${MEM_PERCENT}% RAM" >> /var/log/fivem-mem.log
  # Optional: send to a Discord webhook
  # curl -H "Content-Type: application/json" -d "{\"content\":\"FiveM RAM high: ${MEM_PERCENT}%\"}" $DISCORD_WEBHOOK
fi

Make it executable and schedule it in cron:

sudo chmod +x /usr/local/bin/fivem-mem-check.sh
echo "*/5 * * * * root /usr/local/bin/fivem-mem-check.sh" | sudo tee /etc/cron.d/fivem-mem

Force a simulated crash to see if auto-restart works:

sudo pkill -9 FXServer
sleep 15
sudo systemctl status fivem.service

After 10-15 seconds, the status should show active (running) again, with a short uptime. If it shows failed or inactive, review the service unit.

Check the latest restarts in the systemd log:

sudo journalctl -u fivem.service --since "1 hour ago" | grep -E "Started|Stopped"

Each Started/Stopped pair represents one restart cycle. High frequency indicates the root cause hasn’t been fixed.