feat: add health check endpoint + twice-daily monitoring workflow

New /health-check endpoint verifies:
- Gitea release integrity (manifest version matches tag)
- Duplicate ZIP assets in releases
- EspoCRM API reachability per customer
- Extension load status on each instance
- Stale deployments (logged ok but version not updated)

New n8n workflow runs at 07:00 & 19:00, combines health check
with drift sync, and posts combined report to Mattermost.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
2026-03-30 11:38:36 +00:00
parent f546b7d660
commit 0c77056422
2 changed files with 258 additions and 0 deletions
+130
View File
@@ -439,6 +439,129 @@ def sync_customer_extensions(customer):
}
def health_check_all():
"""Run comprehensive health checks on all extensions and customers."""
issues = []
# 1. Check Gitea release integrity - manifest version matches tag
extensions = query("SELECT name, latest_version FROM extensions WHERE is_active")
for ext in extensions:
module = ext['name']
version = ext['latest_version']
if not version:
continue
tag = f'v{version}'
release = gitea_api('GET', f'/repos/{GITEA_ORG}/{module}/releases/tags/{tag}')
if release.get('error'):
issues.append({'check': 'gitea_release', 'extension': module, 'severity': 'error',
'message': f'Release {tag} not found on Gitea'})
continue
zip_assets = [a for a in release.get('assets', []) if a['name'].endswith('.zip')]
if not zip_assets:
issues.append({'check': 'gitea_release', 'extension': module, 'severity': 'error',
'message': f'No ZIP asset in release {tag}'})
continue
# Check for duplicate ZIP assets
if len(zip_assets) > 1:
issues.append({'check': 'duplicate_assets', 'extension': module, 'severity': 'warning',
'message': f'{len(zip_assets)} ZIP assets in release {tag} — should be 1'})
# Download last ZIP and verify manifest version matches tag
last_zip_url = zip_assets[-1]['browser_download_url']
try:
req = urllib.request.Request(last_zip_url, headers={'Authorization': f'token {GITEA_TOKEN}'})
resp = urllib.request.urlopen(req, context=SSL_CTX, timeout=30)
zip_data = resp.read()
with zipfile.ZipFile(io.BytesIO(zip_data)) as zf:
manifest_files = [n for n in zf.namelist() if n.endswith('manifest.json')]
if manifest_files:
manifest = json.loads(zf.read(manifest_files[0]))
manifest_ver = manifest.get('version', '')
if manifest_ver != version:
issues.append({'check': 'version_mismatch', 'extension': module, 'severity': 'critical',
'message': f'ZIP manifest says {manifest_ver} but release tag is {tag} — wrong ZIP!'})
else:
issues.append({'check': 'manifest_missing', 'extension': module, 'severity': 'error',
'message': f'No manifest.json found inside ZIP for {tag}'})
except Exception as e:
issues.append({'check': 'zip_verify', 'extension': module, 'severity': 'error',
'message': f'Failed to verify ZIP: {str(e)[:150]}'})
# 2. Check EspoCRM API health for each customer
customers = query("""SELECT id, name, espocrm_url, espocrm_api_key
FROM customers WHERE is_active AND espocrm_api_key IS NOT NULL""")
customer_results = []
for c in customers:
espo_url = c['espocrm_url'].rstrip('/')
headers = {'Espo-Authorization': c['espocrm_api_key']}
cr = {'customer': c['name'], 'url': espo_url, 'api_reachable': False, 'extensions_loaded': False}
# API health
try:
req = urllib.request.Request(f'{espo_url}/api/v1/App/user', headers=headers)
resp = urllib.request.urlopen(req, context=SSL_CTX, timeout=15)
data = json.loads(resp.read())
cr['api_reachable'] = True
cr['espo_version'] = data.get('version', 'unknown')
except Exception as e:
issues.append({'check': 'api_health', 'extension': None, 'severity': 'critical',
'message': f'{c["name"]}: EspoCRM API unreachable — {str(e)[:150]}'})
customer_results.append(cr)
continue
# Check extensions are loaded
try:
req = urllib.request.Request(f'{espo_url}/api/v1/Extension?select=name,version&maxSize=200', headers=headers)
resp = urllib.request.urlopen(req, context=SSL_CTX, timeout=15)
data = json.loads(resp.read())
server_exts = {e['name']: e.get('version', '') for e in data.get('list', [])}
cr['extensions_loaded'] = True
cr['extension_count'] = len(server_exts)
# Check installed extensions actually match what platform expects
platform_exts = query("""SELECT e.name, ce.installed_version FROM customer_extensions ce
JOIN extensions e ON ce.extension_id = e.id
WHERE ce.customer_id = %s AND ce.installed_version IS NOT NULL""", (c['id'],))
for pe in platform_exts:
server_ver = server_exts.get(pe['name'])
if server_ver and server_ver != pe['installed_version']:
issues.append({'check': 'install_integrity', 'extension': pe['name'], 'severity': 'warning',
'message': f'{c["name"]}: platform says {pe["installed_version"]} but server has {server_ver}'})
except Exception as e:
issues.append({'check': 'ext_query', 'extension': None, 'severity': 'error',
'message': f'{c["name"]}: failed to query extensions — {str(e)[:150]}'})
customer_results.append(cr)
# 3. Check for stale deployments (status ok but installed_version not updated)
stale = query("""SELECT dl.version as deployed_version, e.name as extension, c.name as customer,
ce.installed_version, dl.started_at
FROM deployment_log dl
JOIN extensions e ON dl.extension_id = e.id
JOIN customers c ON dl.customer_id = c.id
LEFT JOIN customer_extensions ce ON ce.customer_id = dl.customer_id AND ce.extension_id = dl.extension_id
WHERE dl.status = 'ok' AND dl.started_at > NOW() - INTERVAL '24 hours'
AND (ce.installed_version IS NULL OR ce.installed_version != dl.version)
ORDER BY dl.started_at DESC""")
for s in stale:
issues.append({'check': 'stale_deploy', 'extension': s['extension'], 'severity': 'warning',
'message': f'{s["customer"]}: deployed {s["deployed_version"]} but installed shows {s["installed_version"] or "NULL"}'})
critical = len([i for i in issues if i['severity'] == 'critical'])
errors = len([i for i in issues if i['severity'] == 'error'])
warnings = len([i for i in issues if i['severity'] == 'warning'])
return {
'status': 'critical' if critical else 'error' if errors else 'warning' if warnings else 'ok',
'total_issues': len(issues),
'critical': critical, 'errors': errors, 'warnings': warnings,
'issues': issues,
'customers': customer_results
}
def sync_all_customers():
"""Run sync for all active customers."""
customers = query("""SELECT id, name, espocrm_url, espocrm_api_key
@@ -492,6 +615,8 @@ class APIHandler(BaseHTTPRequestHandler):
self._handle_deploy(body)
elif path == '/sync':
self._handle_sync(body)
elif path == '/health-check':
self._handle_health_check()
else:
self._respond(404, {'error': 'Not found'})
@@ -673,6 +798,11 @@ class APIHandler(BaseHTTPRequestHandler):
result = sync_all_customers()
self._respond(200, result)
# --- Health Check ---
def _handle_health_check(self):
result = health_check_all()
self._respond(200, result)
# --- Helpers ---
def _read_body(self):
length = int(self.headers.get('Content-Length', 0))