Fixed Major Agent Deployment Issues

This commit is contained in:
2025-09-27 19:12:29 -06:00
parent 3990a50e8d
commit e4ef065271
4 changed files with 246 additions and 81 deletions

View File

@@ -0,0 +1,120 @@
- hosts: all
gather_facts: false
tasks:
- name: Collect summary via PowerShell
ansible.builtin.shell: |
$ErrorActionPreference = 'Stop'
$hostname = $env:COMPUTERNAME
$username = $env:USERNAME
$domain = $env:USERDOMAIN
$w = Get-ItemProperty 'HKLM:\SOFTWARE\Microsoft\Windows NT\CurrentVersion'
$product = [string]$w.ProductName
$display = [string]$w.DisplayVersion
$build = [string]$w.CurrentBuildNumber
$os = ($product + ' ' + $display).Trim()
$boot = (Get-CimInstance Win32_OperatingSystem).LastBootUpTime
$uptime = [int]((Get-Date) - $boot).TotalSeconds
$epoch = [int](((Get-Date $boot).ToUniversalTime() - [datetime]'1970-01-01').TotalSeconds)
$out = [pscustomobject]@{ hostname=$hostname; os=$os; username=$username; domain=$domain; uptime_sec=$uptime; last_reboot=$epoch }
$out | ConvertTo-Json -Depth 4
register: summary_raw
changed_when: false
- name: Parse summary JSON
ansible.builtin.set_fact:
_summary: "{{ summary_raw.stdout | from_json | default({}, true) }}"
- name: Collect installed software
ansible.builtin.shell: |
$ErrorActionPreference = 'SilentlyContinue'
$paths = @('HKLM:\SOFTWARE\Microsoft\Windows\CurrentVersion\Uninstall\*',
'HKLM:\SOFTWARE\WOW6432Node\Microsoft\Windows\CurrentVersion\Uninstall\*',
'HKCU:\SOFTWARE\Microsoft\Windows\CurrentVersion\Uninstall\*')
$items = foreach ($p in $paths) {
if (Test-Path $p) {
Get-ItemProperty -Path $p | Where-Object { $_.DisplayName } | ForEach-Object {
[pscustomobject]@{ name=[string]$_.DisplayName; version=[string]$_.DisplayVersion }
}
}
}
$items | Sort-Object name -Unique | ConvertTo-Json -Depth 4
register: software_raw
changed_when: false
- name: Parse software JSON
ansible.builtin.set_fact:
_software: "{{ software_raw.stdout | default('[]') | from_json | default([], true) }}"
- name: Collect memory modules
ansible.builtin.shell: |
$ErrorActionPreference = 'SilentlyContinue'
Get-CimInstance Win32_PhysicalMemory | ForEach-Object {
[pscustomobject]@{ slot=$_.BankLabel; speed=[string]$_.Speed; serial=[string]$_.SerialNumber; capacity=[string]$_.Capacity }
} | ConvertTo-Json -Depth 4
register: memory_raw
changed_when: false
- name: Parse memory JSON
ansible.builtin.set_fact:
_memory: "{{ memory_raw.stdout | default('[]') | from_json | default([], true) }}"
- name: Collect storage volumes
ansible.builtin.shell: |
$ErrorActionPreference = 'SilentlyContinue'
Get-CimInstance Win32_LogicalDisk | ForEach-Object {
$total = [double]$_.Size; $free = [double]$_.FreeSpace; $used = $total - $free;
$usage = if ($total -gt 0) { [math]::Round(($used / $total) * 100, 2) } else { 0 }
$type = switch ($_.DriveType) { 2 {'Removable'} 3 {'Fixed Disk'} default {'Unknown'} }
[pscustomobject]@{ drive=$_.DeviceID; disk_type=$type; usage=$usage; total=$total; free=$free; used=$used }
} | ConvertTo-Json -Depth 4
register: storage_raw
changed_when: false
- name: Parse storage JSON
ansible.builtin.set_fact:
_storage: "{{ storage_raw.stdout | default('[]') | from_json | default([], true) }}"
- name: Collect network adapters
ansible.builtin.shell: |
$ErrorActionPreference = 'SilentlyContinue'
try {
$ip = Get-NetIPAddress -AddressFamily IPv4 -ErrorAction Stop | Where-Object { $_.IPAddress -and $_.IPAddress -notmatch '^169\.254\.' -and $_.IPAddress -ne '127.0.0.1' }
$ad = Get-NetAdapter | ForEach-Object { $_ | Select-Object -Property Name, InterfaceAlias, MacAddress }
$map = @{}; foreach($a in $ad){ $map[$a.InterfaceAlias] = $a.MacAddress }
$tmp = @{}
foreach($e in $ip){
$alias = if ($e.InterfaceAlias) { $e.InterfaceAlias } else { 'unknown' }
$item = $tmp[$alias]
if (-not $item) { $item = [pscustomobject]@{ adapter=$alias; ips=@(); mac='' }; $tmp[$alias] = $item }
$item.mac = $map[$alias]
if ($e.IPAddress -and $item.ips -notcontains $e.IPAddress) { $item.ips += $e.IPAddress }
}
$out = $tmp.GetEnumerator() | ForEach-Object { $_.Value }
} catch { $out = @() }
$out | ConvertTo-Json -Depth 4
register: network_raw
changed_when: false
- name: Parse network JSON
ansible.builtin.set_fact:
_network: "{{ network_raw.stdout | default('[]') | from_json | default([], true) }}"
- name: Compose device details structure
ansible.builtin.set_fact:
device_details:
summary: "{{ _summary }}"
software: "{{ _software }}"
memory: "{{ _memory }}"
storage: "{{ _storage }}"
network: "{{ _network }}"
- name: Derive internal IP
ansible.builtin.set_fact:
device_details: "{{ device_details | combine({'summary': (device_details.summary | combine({'internal_ip': ( (_network | map(attribute='ips') | list | first | default([])) | first | default('') ) })) }) }}"
- name: Write device details JSON
ansible.builtin.copy:
content: "{{ device_details | to_nice_json }}"
dest: "{{ output_file }}"
run_once: true

View File

@@ -1,4 +1,5 @@
import os
import sys
import json
import time
import socket
@@ -7,6 +8,7 @@ import subprocess
import shutil
import string
import asyncio
from pathlib import Path
try:
import psutil # type: ignore
@@ -124,6 +126,79 @@ def collect_summary(CONFIG):
return {'hostname': socket.gethostname()}
def _project_root():
try:
# Agent layout: Agent/Borealis/{this_file}; root is two levels up
return os.path.abspath(os.path.join(os.path.dirname(__file__), '..', '..'))
except Exception:
return os.getcwd()
def _run_ansible_audit(ctx) -> dict:
try:
exe_dir = os.path.dirname(sys.executable)
candidate = os.path.join(exe_dir, 'ansible-playbook.exe' if IS_WINDOWS else 'ansible-playbook')
ansible_playbook = candidate if os.path.isfile(candidate) else 'ansible-playbook'
base = os.path.join(_project_root(), 'Logs', 'Agent', 'ansible')
os.makedirs(base, exist_ok=True)
out_path = os.path.join(base, 'audit.json')
# Require an external playbook; look next to this role first, then source tree as fallback
roles_dir = os.path.dirname(__file__)
pb_candidates = [
os.path.join(roles_dir, 'Device_Audit.yml'),
os.path.join(_project_root(), 'Data', 'Agent', 'Roles', 'Device_Audit.yml'),
]
pb_path = next((p for p in pb_candidates if os.path.isfile(p)), None)
if not pb_path:
# Log helpful error and return empty
try:
with open(os.path.join(base, 'ansible.err.log'), 'w', encoding='utf-8', newline='\n') as ef:
ef.write('Device_Audit.yml not found in roles directory.\n')
ef.write('Searched:\n - ' + '\n - '.join(pb_candidates))
except Exception:
pass
return {}
out_posix = Path(out_path).as_posix()
py_interp = Path(sys.executable).as_posix()
env = os.environ.copy()
env.setdefault('PYTHONIOENCODING', 'utf-8')
env.setdefault('ANSIBLE_FORCE_COLOR', '0')
cmd = [
ansible_playbook,
'-i', 'localhost,',
'-c', 'local',
pb_path,
'-e', f'ansible_python_interpreter={py_interp}',
'-e', 'ansible_shell_type=powershell' if IS_WINDOWS else 'ansible_shell_type=sh',
'-e', 'ansible_shell_executable=powershell.exe' if IS_WINDOWS else 'ansible_shell_executable=/bin/sh',
'-e', f'output_file={out_posix}',
]
proc = subprocess.run(cmd, capture_output=True, text=True, timeout=300, env=env)
if proc.returncode != 0:
try:
with open(os.path.join(base, 'ansible.err.log'), 'w', encoding='utf-8', newline='\n') as ef:
ef.write(proc.stdout or '')
ef.write('\n--- STDERR ---\n')
ef.write(proc.stderr or '')
except Exception:
pass
return {}
try:
with open(out_path, 'r', encoding='utf-8') as jf:
details = json.load(jf)
return details if isinstance(details, dict) else {}
except Exception:
return {}
except Exception:
return {}
def _ps_json(cmd: str, timeout: int = 60):
try:
out = subprocess.run(["powershell", "-NoProfile", "-Command", cmd], capture_output=True, text=True, timeout=timeout)
@@ -425,6 +500,9 @@ class Role:
self.ctx = ctx
self._ext_ip = None
self._ext_ip_ts = 0
self._ansible_cache = None
self._ansible_ts = 0
self._last_details = None
try:
# Set OS string once
self.ctx.config.data['agent_operating_system'] = detect_agent_os()
@@ -445,78 +523,35 @@ class Role:
pass
async def _report_loop(self):
interval_sec = 300 # post heartbeat/details every 5 minutes
while True:
try:
details = {
'summary': collect_summary(self.ctx.config),
'software': collect_software(),
'memory': collect_memory(),
'storage': collect_storage(),
'network': collect_network(),
}
# Derive additional summary fields
# Determine audit refresh interval (minutes), default 30
try:
# Internal IP: first IPv4 on first adapter
internal_ip = ''
for a in (details.get('network') or []):
for ip in (a.get('ips') or []):
if ip and not ip.startswith('169.254.') and ip != '127.0.0.1':
internal_ip = ip
break
if internal_ip:
break
details['summary']['internal_ip'] = internal_ip
refresh_min = int(self.ctx.config.data.get('audit_interval_minutes', 30))
except Exception:
pass
try:
details['summary']['device_type'] = detect_device_type()
except Exception:
pass
try:
if psutil:
details['summary']['last_reboot'] = int(psutil.boot_time())
except Exception:
pass
refresh_min = 30
refresh_sec = max(300, refresh_min * 60)
now = time.time()
need_refresh = (not self._last_details) or ((now - self._ansible_ts) > refresh_sec)
if need_refresh:
details = _run_ansible_audit(self.ctx)
if details:
self._last_details = details
self._ansible_ts = now
# Always post the latest available details (possibly cached)
details_to_send = self._last_details or {'summary': collect_summary(self.ctx.config)}
url = (self.ctx.config.data.get('borealis_server_url', 'http://localhost:5000') or '').rstrip('/') + '/api/agent/details'
payload = {
'agent_id': self.ctx.agent_id,
'hostname': details.get('summary', {}).get('hostname', socket.gethostname()),
'details': details,
'hostname': details_to_send.get('summary', {}).get('hostname', socket.gethostname()),
'details': details_to_send,
}
if aiohttp is not None:
async with aiohttp.ClientSession() as session:
# External IP: refresh at most every 30 minutes
try:
now = time.time()
if (now - self._ext_ip_ts) > 1800:
# Try ipify JSON; fallback to plain-text ifconfig.me
ok = False
try:
async with session.get('https://api.ipify.org?format=json', timeout=8) as resp:
if resp.status == 200:
j = await resp.json()
self._ext_ip = (j.get('ip') or '').strip()
self._ext_ip_ts = now
ok = True
except Exception:
pass
if not ok:
try:
async with session.get('https://ifconfig.me/ip', timeout=8) as resp2:
if resp2.status == 200:
t = (await resp2.text()) or ''
t = t.strip()
if t:
self._ext_ip = t
self._ext_ip_ts = now
ok = True
except Exception:
pass
if self._ext_ip:
details['summary']['external_ip'] = self._ext_ip
except Exception:
pass
await session.post(url, json=payload, timeout=10)
except Exception:
pass
await asyncio.sleep(300)
await asyncio.sleep(interval_sec)