SLU Cluster Management (infra-ansible)

This repository contains the Ansible orchestration for the Open Source SLU server cluster. It manages the configuration, security, and deployment for oss01 (Master) and oss02 (Worker).

---

Quick Start

Provision the entire infrastructure from scratch:

ansible-playbook -i inventory/hosts.ini site.yaml -K

Run Specific Layers

System tools only:

ansible-playbook site.yaml --tags base -K

Kubernetes and networking only:

ansible-playbook -i inventory/hosts.ini site.yaml --tags base -K

---

Cluster Sanity Checks

Run these commands to verify infrastructure health.
If all checks pass, the cluster is healthy.

---

1. Node Status

Verify master and worker communication:

k get nodes -o wide

Expected result:

STATUS: Ready

Both oss01 and oss02 should show Ready.

---

2. Pod Distribution

Ensure workloads are balanced across nodes (verifies Calico mesh networking):

k get pods -A -o wide | grep -v 'kube-system'

Expected result:

Pods appear on both nodes under the NODE column.

---

3. Service Connectivity

Verify NodePort exposure and application response:

# Check if the OS is listening on the NodePort
sudo netstat -tulpn | grep 30080

# Test the HTTP response
curl -I http://localhost:30080

Expected result:

HTTP/1.1 200 OK

---

4. Application Logs (Prisma / Database)

Confirm backend-to-database connectivity:

k logs deploy/coredesk-app --tail=50

Expected result:

prisma:client libraryStarted: true

There should be no database connection errors.

---

Troubleshooting

Symptom	Probable Cause	Fix
Apt cache lock	Broken or third-party repo issue	Run ansible-playbook site.yaml --tags base -K
Node NotReady	Calico / CNI handshake failure	Run ansible-playbook site.yaml --tags k8s_prep -K
Connection Refused	Firewall or NodePort mismatch	Verify port with k get svc

---

Security Note

Sensitive variables in group_vars/all.yml are encrypted using Ansible Vault.

To run playbooks that require secrets:

ansible-playbook site.yaml -K --ask-vault-pass

---

Repository Structure

infra-ansible/
├── inventory/
│   └── hosts.ini
├── group_vars/
│   └── all.yml (vault encrypted)
├── roles/
├── site.yaml
└── README.md

---

Cluster Health Definition

The cluster is considered healthy when:

• Both nodes are Ready
• Pods are distributed across nodes
• NodePort responds with 200 OK
• Prisma initializes without database errors
• No Calico or CNI failures are present

---

Maintained by the Open Source SLU Infrastructure & Operations Team