500, 502, 503 and 504 errors, resource-limit warnings, timeouts and slow server response. We read the logs, find what’s failing (PHP, the database, hosting limits or a proxy) and fix the cause.
Tell us what is happening and we will reply with a clear next step.
Server troubleshooting means finding out why the machine and software that run your website (web server, PHP, database and hosting limits) are failing to serve pages, then fixing the configuration, the code that’s overloading it, or the hosting setup itself.
The fastest way to narrow a problem down is to look at where a request stops. When someone visits your site, the request passes through DNS, often a CDN or proxy, the web server, PHP, and the database. Each stage fails in its own way, and the error code tells you roughly where.
| Code | What it means | Common causes |
|---|---|---|
| 500 Internal Server Error | The server hit an error it couldn’t handle | PHP fatal error, bad .htaccess rule, wrong file permissions |
| 502 Bad Gateway | A proxy or CDN got an invalid response from your server | PHP process crashed, server overloaded, firewall blocking the CDN |
| 503 Service Unavailable | The server is temporarily refusing requests | Resource limits reached, maintenance mode, PHP workers exhausted |
| 504 Gateway Timeout | The server took too long to answer | Slow database queries, long-running PHP, external API calls hanging |
| 508 Resource Limit Reached | Shared-hosting account hit its limits | Traffic spikes, bots, heavy plugins, runaway cron jobs |
| 403 Forbidden | Access refused | Security rules, file permissions, a firewall or .htaccess block |
| 429 Too Many Requests | Rate limiting kicked in | Aggressive bots, misconfigured security tools |
| 520–524 | Cloudflare-specific errors between Cloudflare and your server | Origin down, timeouts, SSL mismatch between Cloudflare and host |
These are starting points. The server’s error logs tell us which applies to your site.
Shared and managed hosting plans limit how much CPU, memory, simultaneous processes and file count (inodes) your account can use. When a site reaches those limits, pages slow down and then fail with 503 or 508 errors, often at the busiest times.
Usually something specific is using more than it should: a plugin running heavy database queries on every page, a search feature hit by bots, uncached pages under load, WP-Cron firing on every visit, or an attack hammering the login page. We identify what is consuming resources before recommending a bigger plan. Upgrading hosting hides the problem; it doesn’t always fix it.
Access logs, slow-query logs and process lists show which pages, bots or tasks use the resources.
Block or rate-limit abusive bots and login attacks before they reach PHP.
Page, object and browser caching so most visits never touch the database.
Many server errors come down to configuration that no longer matches what the site needs:
.htaccess; Nginx uses its own configuration. A single wrong redirect or security rule can take a site down or create a redirect loop.We change one thing at a time, note the previous value, and test after each change, so anything can be undone.
“Error establishing a database connection” means WordPress couldn’t reach its database: wrong credentials after a move, a database server that’s down or overloaded, or corrupted tables.
Slower, quieter database problems are more common: tables bloated with old revisions and expired transients, large amounts of autoloaded options loaded on every request, missing indexes, or one plugin running an expensive query on every page. We find them with slow-query logs and query monitoring, then clean up and optimise. We always take a backup first.
WordPress runs scheduled tasks (publishing posts, sending emails, running backups, clearing caches) through WP-Cron, which is triggered by visits rather than a real clock. On quiet sites tasks run late; on busy sites they run too often; and a stuck task can loop and overload the server.
On most production sites we switch to a real server cron job that runs WordPress tasks at a fixed interval, and we check for plugins that schedule more work than they need.
Then you get a short write-up: what was wrong, what we changed, and anything you should know about your hosting plan.
Sometimes the honest answer is that the hosting plan doesn’t fit the site anymore: a busy WooCommerce store on entry-level shared hosting, a host that can’t offer a supported PHP version, or one that won’t give you access to logs. We’ll tell you when that’s the case, explain the options in plain terms, and if you decide to move, handle it as a planned hosting migration.
We don’t receive commissions from hosting companies, so recommendations are based on what your site needs.
cPanel, Plesk, a managed host’s dashboard, or SSH on a VPS.
If you use Cloudflare or similar, so we can see both ends.
Times, pages affected and any pattern, such as busy periods.
Plan name and limits, if you know them.
Yes, as well as shared and managed WordPress hosting. On VPS and dedicated servers we can work at the operating-system and web-server level if you give us SSH access.
Sometimes it is, sometimes it’s the hosting configuration, and often it’s both. Logs and resource graphs settle the question, and we’ll show you the evidence either way.
Usually not. Most investigation is read-only. If a change needs a short interruption, we’ll schedule it for your quiet hours and tell you first.
Yes. 52x errors usually mean Cloudflare can’t get a good response from your server, which can be an origin problem, an SSL mode mismatch or a firewall blocking Cloudflare.
We recommend based on your site’s needs and budget, and we don’t take referral commissions.
info@webexpert.solutions · +92 345 7723858 · Faisalabad, Pakistan · Serving the UAE, US, UK & Canada