homelab-optimized/docs/admin/monitoring-update-seattle-2026-02.md

Seattle Machine Monitoring Update

Summary

Successfully updated the homelab monitoring system to replace the decommissioned VMI (100.99.156.20) with the reprovisioned Seattle machine (100.82.197.124).

Changes Made

1. Prometheus Configuration Update

File: /home/homelab/docker/monitoring/prometheus/prometheus.yml

Before:

- job_name: "vmi2076105-node"
  static_configs:
    - targets: ["100.99.156.20:9100"]

After:

- job_name: "seattle-node"
  static_configs:
    - targets: ["100.82.197.124:9100"]

2. Seattle Machine Configuration

Node Exporter Installation

• Node exporter was already running on the Seattle machine
• Service status: active (running) on port 9100
• Binary location: /usr/local/bin/node_exporter

Firewall Configuration

Added UFW rule to allow Tailscale network access:

sudo ufw allow from 100.64.0.0/10 to any port 9100 comment 'Allow Tailscale to node_exporter'

SSH Access

• Accessible via ssh seattle-tailscale (configured in SSH config)
• Tailscale IP: 100.82.197.124
• Standard SSH key authentication

3. Monitoring Verification

Prometheus Targets Status

All monitoring targets are now healthy:

• prometheus: localhost:9090 ✅ UP
• alertmanager: alertmanager:9093 ✅ UP
• node-exporter: localhost:9100 ✅ UP
• calypso-node: 100.75.252.64:9100 ✅ UP
• seattle-node: 100.82.197.124:9100 ✅ UP
• proxmox-node: 100.87.12.28:9100 ✅ UP

Metrics Collection

• Seattle machine metrics are being successfully scraped
• CPU, memory, disk, and network metrics available
• Historical data collection started immediately after configuration

Technical Details

Network Configuration

• Tailscale Network: 100.64.0.0/10
• Seattle IP: 100.82.197.124
• Monitoring Port: 9100 (node_exporter)
• Protocol: HTTP (internal network)

Service Architecture

Prometheus (homelab) → Tailscale Network → Seattle Machine:9100 (node_exporter)

Configuration Files Updated

1. /home/homelab/docker/monitoring/prometheus/prometheus.yml - Production config
2. /home/homelab/organized/repos/homelab/prometheus/prometheus.yml - Repository config
3. Fixed YAML indentation issues for alertmanager targets

Verification Steps Completed

1. ✅ SSH connectivity to Seattle machine
2. ✅ Node exporter service running and accessible
3. ✅ Firewall rules configured for Tailscale access
4. ✅ Prometheus configuration updated and reloaded
5. ✅ Target health verification (UP status)
6. ✅ Metrics scraping confirmed
7. ✅ Repository configuration synchronized
8. ✅ Git commit with detailed change log

Monitoring Capabilities

The Seattle machine now provides the following metrics:

• System: CPU usage, load average, uptime
• Memory: Total, available, used, cached
• Disk: Usage, I/O statistics, filesystem metrics
• Network: Interface statistics, traffic counters
• Process: Running processes, file descriptors

Alert Coverage

The Seattle machine is now covered by all existing alert rules:

• InstanceDown: Triggers if node_exporter becomes unavailable
• HighCPUUsage: Alerts when CPU usage > 80% for 2+ minutes
• HighMemoryUsage: Alerts when memory usage > 90% for 2+ minutes
• DiskSpaceLow: Alerts when root filesystem < 10% free space

Next Steps

1. Monitor Performance: Watch Seattle machine metrics for baseline establishment
2. Alert Tuning: Adjust thresholds if needed based on Seattle machine characteristics
3. Documentation: This update is documented in the homelab repository
4. Backup Verification: Ensure Seattle machine is included in backup monitoring

Rollback Plan

If issues arise, the configuration can be quickly reverted:

# Revert Prometheus config
cd /home/homelab/docker/monitoring
git checkout HEAD~1 prometheus/prometheus.yml
docker compose restart prometheus

Contact Information

• Updated By: OpenHands Agent
• Date: February 15, 2026
• Commit: fee90008 - "Update monitoring: Replace VMI with Seattle machine"
• Repository: homelab.git

---

Status: ✅ COMPLETED SUCCESSFULLY Monitoring: ✅ ACTIVE AND HEALTHY Documentation: ✅ UPDATED