Files
dashcaddy/dashcaddy-api/src/utilities/startup-validator.js
T
Krystie e99413150e
CI / Test & Lint (push) Canceled after 0s
CI / Security audit (push) Canceled after 0s
[glm-grade=B] fix: dead dashboard WS, corrupted error.log, auth-polling storm, re-auth freeze + CVE bumps
Adversarial audit 2026-08-16 (GLM-5.3 delegate, 2 rounds, 141 tool calls):

P0-1: Dashboard WebSocket (/api/v1/ws) dead on EVERY boot since DC-076.
server.js passed module exports (DependencyManager class, {AutoRestartManager}
namespace, SSLMonitor class) instead of createApp()'s live instances — first
.on() threw ERR_INVALID_ARG_TYPE, catch swallowed it. Fix: app.locals.ctx
exposed in src/app.js; server.js passes all 8 real EventEmitter instances.

P0-2: error.log corrupted since 2026-07-14. errorMiddleware called
logError(FILE, SIZE, path, err, meta) — 5 args into a 3-arg wrapper —
logging 'Error: 5242880' garbage every ~60s and DISCARDING the real error
object. Fix: correct 3-arg call + legacy-shape guard in logErrorWrapper +
~74 log.error sites swept to pass real error objects (AST-verified scope-
safe 71/71, 29/29 modules load clean).

P0-3: auth-polling storm (stranded grade=B commit never landed in prod):
401/403 behind TOTP gate hammered /api/v1/services/status + SSE reconnect
every 2-8s, with misleading direct-probe fallback marking services 'up'.
Fix landed + B-round MEDIUM follow-up: TOTP re-auth success now clears
_dcAuthLost, resumes SSE (new _sseResume clears the latch), and refreshes.

Also: eslintignore static-sites/ (33→0 errors); nodemailer 8→9.0.5 and
sharp 0.33→0.35.3 (3 high CVEs killed; jest green on new majors);
dockerode@5/uuid deferred (semver-major, Docker API surface).

Verification: 80/80 suites, 1837/1837 tests; ESLint 0 errors/743 warnings;
node --check all changed files; bundles rebuilt + SW cache bumped.

Judges: Codex quota-dead until Aug 19 (verified live) — GLM adversarial
delegate per operator directive 2026-08-07. Round 1: 98-call mechanical
verification (timed out pre-verdict). Round 2 (this grade): B, one MEDIUM
(re-auth freeze) — fixed in this commit as prescribed.
2026-08-16 04:18:07 -07:00

236 lines
8.0 KiB
JavaScript

/**
* Startup Validation Module
* Extracts startup configuration validation and health checker sync from server.js
* (Phase 3 refactoring)
*/
const fs = require('fs');
const fsp = require('fs').promises;
const path = require('path');
const http = require('http');
const https = require('https');
const { exists, isAccessible } = require('./fs-helpers');
const { resolveServiceUrl } = require('./url-resolver');
/**
* Validate startup configuration and environment.
* Fail fast with clear error messages if critical issues are found.
*
* @param {Object} deps
* @param {Function} deps.log - structured logger
* @param {string} deps.CADDYFILE_PATH
* @param {string} deps.SERVICES_FILE
* @param {string} deps.CONFIG_FILE
* @param {string} deps.CADDY_ADMIN_URL
* @param {number} deps.PORT
*/
async function validateStartupConfig({ log, CADDYFILE_PATH, SERVICES_FILE, CONFIG_FILE, CADDY_ADMIN_URL, PORT }) {
const errors = [];
const warnings = [];
log.info('startup', 'Validating startup configuration...');
// 1. Check if Caddyfile exists and is writable
try {
if (await exists(CADDYFILE_PATH)) {
if (!(await isAccessible(CADDYFILE_PATH, fs.constants.R_OK | fs.constants.W_OK))) {
errors.push(`Caddyfile is not readable/writable: ${CADDYFILE_PATH}`);
} else {
log.info('startup', 'Caddyfile is accessible', { path: CADDYFILE_PATH });
}
} else {
warnings.push(`Caddyfile does not exist: ${CADDYFILE_PATH} (will be created if needed)`);
}
} catch (error) {
errors.push(`Caddyfile is not readable/writable: ${CADDYFILE_PATH}`);
}
// 2. Check if services file exists or can be created
const servicesDir = path.dirname(SERVICES_FILE);
try {
if (await exists(SERVICES_FILE)) {
// Validate it's valid JSON
try {
const content = await fsp.readFile(SERVICES_FILE, 'utf8');
JSON.parse(content);
log.info('startup', 'Services file is valid JSON', { path: SERVICES_FILE });
} catch (parseError) {
errors.push(`Services file exists but contains invalid JSON: ${SERVICES_FILE}`);
}
} else {
// Check if parent directory exists and is writable
if (await exists(servicesDir)) {
if (!(await isAccessible(servicesDir, fs.constants.W_OK))) {
errors.push(`Cannot access services file or directory: ${SERVICES_FILE}`);
} else {
log.info('startup', 'Services file directory is writable', { path: servicesDir });
}
} else {
errors.push(`Services file directory does not exist: ${servicesDir}`);
}
}
} catch (error) {
errors.push(`Cannot access services file or directory: ${SERVICES_FILE}`);
}
// 3. Check if port is available
// CRITICAL: listen() and close() are async. If we fire-and-forget both
// (the old code), the kernel hasn't released the port by the time
// app.listen(PORT) runs in server.js → EADDRINUSE → crash loop.
// Await both via Promises so the port is truly free before we return.
const net = require('net');
const portCheckServer = net.createServer();
try {
await new Promise((resolve, reject) => {
portCheckServer.once('error', reject);
portCheckServer.listen(PORT, '0.0.0.0', () => {
portCheckServer.close(() => {
portCheckServer.removeListener('error', reject);
resolve();
});
});
});
log.info('startup', `Port ${PORT} is available`);
} catch (error) {
errors.push(`Port ${PORT} is already in use or cannot be bound`);
}
// 4. Check if config file is valid JSON (if exists)
try {
if (await exists(CONFIG_FILE)) {
const content = await fsp.readFile(CONFIG_FILE, 'utf8');
JSON.parse(content);
log.info('startup', 'Config file is valid JSON', { path: CONFIG_FILE });
} else {
log.warn('startup', 'Config file does not exist (will use defaults)', { path: CONFIG_FILE });
}
} catch (error) {
errors.push(`Config file exists but contains invalid JSON: ${CONFIG_FILE}`);
}
// 5. Check Caddy admin API reachability (warning only, not critical)
const checkCaddyAdmin = () => {
return new Promise((resolve) => {
const urlObj = new URL(CADDY_ADMIN_URL);
const client = urlObj.protocol === 'https:' ? https : http;
const req = client.request({
hostname: urlObj.hostname,
port: urlObj.port,
path: '/config/',
method: 'GET',
timeout: 2000
}, (res) => {
resolve(res.statusCode >= 200 && res.statusCode < 500);
});
req.on('error', () => resolve(false));
req.on('timeout', () => {
req.destroy();
resolve(false);
});
req.end();
});
};
// Run async Caddy check (don't block startup)
checkCaddyAdmin().then(isReachable => {
if (isReachable) {
log.info('startup', 'Caddy admin API is reachable', { url: CADDY_ADMIN_URL });
} else {
log.warn('startup', 'Caddy admin API is not reachable (may start later)', { url: CADDY_ADMIN_URL });
}
});
// Print warnings
if (warnings.length > 0) {
warnings.forEach(warning => log.warn('startup', warning));
}
// Fail fast if there are critical errors
if (errors.length > 0) {
errors.forEach(err => log.error('startup', err));
log.error('startup', 'Cannot start server — fix the above errors and try again');
process.exit(1);
}
log.info('startup', 'Startup configuration validation passed');
}
/**
* Full-sync health checker from services.json + top-card services.
* Adds missing, updates changed URLs, removes deleted services.
*
* @param {Object} deps
* @param {Function} deps.log - structured logger
* @param {string} deps.SERVICES_FILE
* @param {Object} deps.servicesStateManager - StateManager instance
* @param {Object} deps.healthChecker - health checker module
* @param {Function} deps.buildServiceUrl - canonical URL builder
* @param {Object} deps.siteConfig - site config (dnsServers, etc.)
* @param {Object} deps.APP - app constants (USER_AGENTS)
*/
async function syncHealthCheckerServices({ log, SERVICES_FILE, servicesStateManager, healthChecker, buildServiceUrl, siteConfig, APP }) {
try {
const topCardServices = [
{ id: 'internet', name: 'Internet' },
];
// Dynamically add all configured DNS servers from config
for (const [id, info] of Object.entries(siteConfig?.dnsServers || {})) {
topCardServices.push({ id, name: info.name || id.toUpperCase() });
}
let appServices = [];
if (await exists(SERVICES_FILE)) {
const data = await servicesStateManager.read();
appServices = Array.isArray(data) ? data : data.services || [];
}
const allServices = [...topCardServices, ...appServices];
const desiredIds = new Set();
let added = 0, updated = 0, removed = 0;
for (const svc of allServices) {
const id = svc.id || svc.name?.toLowerCase();
if (!id) continue;
desiredIds.add(id);
const url = resolveServiceUrl(id, svc, siteConfig, buildServiceUrl);
const existing = healthChecker.config.services?.[id];
if (!existing) {
healthChecker.configureService(id, {
name: svc.name || id,
url,
method: 'HEAD',
timeout: 5000,
expectedStatusCodes: [200, 201, 204, 301, 302, 303, 307, 308, 401, 403],
headers: { 'User-Agent': APP.USER_AGENTS.HEALTH },
});
added++;
} else if (existing.url !== url) {
healthChecker.configureService(id, { ...existing, url });
updated++;
}
}
// Remove services no longer in the desired set
const configuredIds = Object.keys(healthChecker.config.services || {});
for (const id of configuredIds) {
if (!desiredIds.has(id)) {
healthChecker.removeService(id);
removed++;
}
}
if (added > 0 || updated > 0 || removed > 0) {
log.info('health', 'Health checker synced', { added, updated, removed });
}
} catch (error) {
log.error('health', error, null, { note: 'Error syncing health checker' });
}
}
module.exports = { validateStartupConfig, syncHealthCheckerServices };