Edge Infrastructure Management

Use Expanso to manage, monitor, and automate distributed edge infrastructure. Execute commands, collect logs, deploy workloads, and maintain infrastructure across thousands of edge locations—all through declarative pipelines.

Why Manage Infrastructure with Expanso?

Traditional edge infrastructure challenges:

• Manual SSH to hundreds of locations doesn't scale
• Logs scattered across distributed sites
• Difficult to enforce consistency
• No centralized operational visibility

Expanso advantages:

• Declarative operations as pipelines
• Automated execution across all edge nodes
• Centralized log collection and monitoring
• Works offline with automatic retry
• Unified control plane for all locations

---

Getting Started

Choose your infrastructure platform to get started:

Kubernetes, K3s & OpenShift

📋 Collect K3s Logs

Stream kubectl logs from all pods and namespaces to centralized storage (S3, Elasticsearch, etc.)

⚙️ Single-Node OpenShift

Manage SNO deployments at the edge. Collect logs, monitor health, handle offline scenarios.

🚀 Deploy Kubernetes Manifests

GitOps-style deployments: apply manifests from S3/Git to edge K3s clusters automatically

Docker & Containers

🐳 Docker Compose Management

Deploy stacks, collect logs, monitor containers. Auto-restart failed services.

🔐 Secrets Management

Provision and mount bootstrap tokens & credentials securely before installation.

---

Common Use Cases

Log Aggregation:

• Collect K3s/Kubernetes logs → S3/Elasticsearch
• Stream Docker container logs to centralized platform
• Parse and enrich logs with node/location metadata

Automated Deployments:

• GitOps: Deploy manifests from Git/S3 to edge clusters
• Manage docker-compose stacks remotely
• Roll out updates across distributed infrastructure

Health Monitoring:

• Check Kubernetes node and pod health
• Monitor Docker daemon and container status
• Alert on infrastructure issues automatically

Maintenance Automation:

• Scheduled backups of cluster state
• Prune unused Docker images/containers
• Clean up old logs and temporary files

---

Quick Example: K3s Log Collection

Stream all K3s logs to S3 in 3 steps:

input:
  subprocess:
    name: kubectl
    args: [logs, --all-containers, --prefix, --follow, --all-namespaces]
    codec: lines
    restart_on_exit: true

pipeline:
  processors:
    - mapping: |
        root = this
        root.node_id = env("NODE_ID")
        root.timestamp = now()

output:
  aws_s3:
    bucket: edge-k3s-logs
    path: 'logs/${! env("NODE_ID") }/${! timestamp_unix() }.jsonl'
    batching:
      count: 1000
      period: 1m

See full example →

---

Architecture Patterns

Hub-and-Spoke Log Collection

Edge Sites (K3s/Docker) 
  → Expanso Agent 
  → Central Storage (S3/Elasticsearch)
  → Analysis/Dashboards

GitOps Deployments

Git/S3 (manifests)
  → Expanso watches for changes
  → kubectl apply to edge clusters
  → Report status to control plane

Self-Healing Infrastructure

Health Check (every 30s)
  → Detect issues (failed pods/containers)
  → Auto-remediation (restart, redeploy)
  → Alert + log results

---

Best Practices

1. Add node identifiers to all outputs for filtering
2. Batch cloud writes to reduce API calls and costs
3. Use subprocess for long-running commands (kubectl logs -f)
4. Handle command failures gracefully with error handling
5. Set resource limits to prevent memory issues

---

Next Steps

• Choose a guide above to get started
• Command Processor: Execute commands per message
• Subprocess Input: Stream command output
• Error Handling: Handle infrastructure failures