ilia c3e6caf9e8
All checks were successful
CI / skip-ci-check (push) Successful in 1m18s
CI / lint-and-test (push) Successful in 1m23s
CI / ansible-validation (push) Successful in 3m2s
CI / secret-scanning (push) Successful in 1m19s
CI / dependency-scan (push) Successful in 1m24s
CI / sast-scan (push) Successful in 2m32s
CI / license-check (push) Successful in 1m23s
CI / vault-check (push) Successful in 2m22s
CI / playbook-test (push) Successful in 2m25s
CI / container-scan (push) Successful in 1m51s
CI / sonar-analysis (push) Successful in 2m32s
CI / workflow-summary (push) Successful in 1m17s
refactor-servers-workstations-shell-monitoring (#4)
### Summary

This PR refactors the playbook layout to reduce duplication and make host intent clearer (servers vs workstations), splits monitoring by host type, and restores full Zsh setup for developers while keeping servers aliases-only.

### Key changes

- **New playbooks**
  - `playbooks/servers.yml`: baseline for server-class hosts (no desktop apps)
  - `playbooks/workstations.yml`: baseline for dev/desktop/local + **desktop apps only on `desktop` group**

- **Monitoring split**
  - `roles/monitoring_server`: server monitoring + intrusion prevention (includes `fail2ban`, sysstat)
  - `roles/monitoring_desktop`: desktop-oriented monitoring tooling
  - Updated playbooks to use the correct monitoring role per host type

- **Shell role: server-safe + developer-friendly**
  - `roles/shell` now supports two modes:
    - `shell_mode: minimal` (default): aliases-only, does not overwrite `.zshrc`
    - `shell_mode: full`: installs Oh My Zsh + Powerlevel10k + plugins and deploys a managed `.zshrc`
  - `playbooks/development.yml` and `playbooks/workstations.yml` use `shell_mode: full`
  - `playbooks/servers.yml` remains **aliases-only**

- **Applications**
  - Applications role runs only on `desktop` group (via `workstations.yml`)
  - Removed Brave installs/repo management
  - Added **CopyQ** to desktop apps (`applications_desktop_packages`)

- **Docs + architecture**
  - Added canonical doc tree under `project-docs/` (overview/architecture/standards/workflow/decisions)
  - Consolidated architecture docs: `docs/reference/architecture.md` is now a pointer to `project-docs/architecture.md`
  - Fixed broken doc links by adding the missing referenced pages under `docs/`

### Behavior changes (important)

- Desktop GUI apps install **only** on the `desktop` inventory group (not on servers, not on dev VMs unless they are in `desktop`).
- Dev/workstation Zsh is now provisioned in **full mode** (managed `.zshrc` + p10k).

### How to test (local CI parity)

```bash
make test
npm test
```

Optional dry runs (interactive sudo may be required):

```bash
make check
make check-local
```

### Rollout guidance

- Apply to a single host first:
  - Workstations: `make workstations HOST=<devhost>`
  - Servers: `make servers HOST=<serverhost>`
- Then expand to group runs.

Reviewed-on: #4
2026-01-01 22:11:24 -05:00

149 lines
4.0 KiB
YAML

---
- name: Install monitoring packages (server)
ansible.builtin.apt:
name:
# System monitoring
- htop
- iotop
- nethogs
- iftop
- ncdu
- dstat
# Log monitoring / security
- logwatch
- fail2ban
# Network monitoring
- nmap
- tcpdump
# Performance monitoring
- sysstat
- atop
state: present
- name: Check if btop is available in apt
ansible.builtin.command: apt-cache policy btop
register: monitoring_server_btop_apt_check
changed_when: false
failed_when: false
when: monitoring_server_install_btop | bool
- name: Install btop from apt if available (Debian 12+)
ansible.builtin.apt:
name: btop
state: present
update_cache: false
when:
- monitoring_server_install_btop | bool
- monitoring_server_btop_apt_check.rc == 0
- "'Candidate:' in monitoring_server_btop_apt_check.stdout"
- "'(none)' not in monitoring_server_btop_apt_check.stdout"
failed_when: false
- name: Install btop from binary if apt not available
when:
- monitoring_server_install_btop | bool
- monitoring_server_btop_apt_check.rc != 0 or "(none)" in monitoring_server_btop_apt_check.stdout
block:
- name: Download btop binary
ansible.builtin.get_url:
url: https://github.com/aristocratos/btop/releases/latest/download/btop-x86_64-linux-musl.tbz
dest: /tmp/btop.tbz
mode: '0644'
failed_when: false
- name: Extract btop
ansible.builtin.unarchive:
src: /tmp/btop.tbz
dest: /tmp/
remote_src: true
failed_when: false
- name: Install btop binary
ansible.builtin.copy:
src: /tmp/btop/bin/btop
dest: /usr/local/bin/btop
mode: '0755'
remote_src: true
failed_when: false
- name: Clean up btop download
ansible.builtin.file:
path: "{{ item }}"
state: absent
loop:
- /tmp/btop.tbz
- /tmp/btop
failed_when: false
- name: Configure fail2ban
ansible.builtin.template:
src: jail.local.j2
dest: /etc/fail2ban/jail.local
mode: '0644'
notify: restart fail2ban
- name: Enable sysstat data collection
ansible.builtin.lineinfile:
path: /etc/default/sysstat
regexp: '^ENABLED='
line: 'ENABLED="true"'
notify: restart sysstat
when: monitoring_server_enable_sysstat | bool
- name: Create monitoring scripts directory
ansible.builtin.file:
path: /usr/local/bin/monitoring
state: directory
mode: '0755'
when: monitoring_server_create_scripts | bool
- name: Deploy system monitoring script
ansible.builtin.copy:
content: |
#!/bin/bash
# System monitoring dashboard
echo "=== System Overview ==="
echo "Hostname: $(hostname)"
echo "Uptime: $(uptime -p)"
echo "Load: $(uptime | awk -F'load average:' '{print $2}')"
echo ""
echo "=== Memory ==="
free -h
echo ""
echo "=== Disk Usage ==="
df -h / /home 2>/dev/null | grep -v tmpfs
echo ""
echo "=== Top Processes ==="
ps aux --sort=-%cpu | head -6
echo ""
echo "=== Network Connections ==="
ss -tuln | head -10
echo ""
if command -v tailscale >/dev/null; then
echo "=== Tailscale Status ==="
tailscale status --peers=false 2>/dev/null || echo "Not connected"
fi
dest: /usr/local/bin/monitoring/sysinfo
mode: '0755'
when: monitoring_server_create_scripts | bool
- name: Deploy network monitoring script
ansible.builtin.copy:
content: |
#!/bin/bash
# Network monitoring script
echo "=== Network Interface Status ==="
ip addr show | grep -E "(inet |state )" | grep -v 127.0.0.1
echo ""
echo "=== Route Table ==="
ip route show
echo ""
echo "=== DNS Configuration ==="
cat /etc/resolv.conf | grep nameserver
echo ""
echo "=== Open Ports ==="
ss -tuln | grep LISTEN | sort
dest: /usr/local/bin/monitoring/netinfo
mode: '0755'
when: monitoring_server_create_scripts | bool