Server Monitoring

Keeping game servers healthy is essential for a good player experience. This page covers the tools and techniques for monitoring your PlayRen game servers.

Health Check Endpoint

Every pool agent exposes a health endpoint:

GET https://gs01.play.ren.bd/api/health

Response:

{
  "status": "ok",
  "containers": 7,
  "cpu_percent": 42.5,
  "memory_percent": 58.3,
  "disk_percent": 35.2
}

Fields

Automated Health Checks

The platform server polls each registered game server's health endpoint periodically. Servers that fail health checks are temporarily removed from the scheduler's rotation until they recover.

You can also set up external monitoring (e.g., UptimeRobot, Pingdom) to alert you if a server goes down.

Container Metrics

Listing Active Containers

GET https://gs01.play.ren.bd/api/containers
Authorization: Bearer {agent-secret}

Returns a list of all running game containers with their resource usage.

Per-Container Stats

When a container is stopped, the pool agent collects its resource usage:

• CPU time — total CPU seconds consumed.
• Peak memory — maximum memory usage during the session.
• Duration — how long the container ran.

These metrics are sent back to the platform and used to update each game's resource profile (exponential moving average), which informs the scheduler's capacity planning.

Docker Stats

For real-time container monitoring, use Docker's built-in stats:

# All containers
docker stats
 
# Specific container
docker stats container-name

Disk Usage

Game cache is the primary consumer of disk space. Monitor it with:

# Overall disk usage
df -h /srv/games
 
# Per-game cache sizes
du -sh /srv/games/*

Cache Management

If disk usage gets high, you can manually clear cached games that haven't been used recently:

# Check which games are cached
ls -la /srv/games/
 
# Remove a specific game's cache
rm -rf /srv/games/game-slug/

The platform will re-download the game on the next launch request. The GameCacheEntry database record is updated automatically.

Future versions will include automatic LRU cache eviction.

Logs

Pool Agent Logs

The pool agent runs as a systemd service. View logs with:

# Follow live logs
journalctl -u xgame9-agent -f
 
# Last 100 lines
journalctl -u xgame9-agent -n 100
 
# Logs from today
journalctl -u xgame9-agent --since today

Docker Container Logs

View logs from a specific game container:

docker logs container-name
docker logs -f container-name  # Follow

Nginx Logs

# Access log
tail -f /var/log/nginx/access.log
 
# Error log
tail -f /var/log/nginx/error.log

Common Alerts to Set Up

Consider monitoring for these conditions:

Server Resource Planning

Track these metrics over time to plan capacity:

• Peak concurrent sessions — highest number of simultaneous containers.
• Average session duration — how long players typically play.
• Cache hit rate — percentage of launches that find the game already cached.
• Launch latency — time from launch request to stream connection.

Next Steps

• Configuration — Adjust agent settings and limits.
• Overview — Review the overall architecture.