Please wait while we load the content.
VPS reviews · Tutorials · Technical articles
246 articles about hosting reviews, deployment and buying advice.
If your VPS website keeps redirecting and the browser reports too many redirects, it's usually Cloudflare SSL mode, Nginx forced HTTPS, reverse proxy headers, or the app's backend address fighting each other. Check layer by layer along the chain—don't just reinstall the certificate right away.
When a VPS website shows 413 Request Entity Too Large while uploading images, backup archives, or large files, it is usually because the request body size limits of Nginx, PHP-FPM, the application framework, or the CDN are inconsistent. This article troubleshoots client_max_body_size, upload_max_filesize, post_max_size, reverse proxy, and Cloudflare limits layer by layer.
After a VPS reboot, your website, Docker containers, Nginx, or MySQL don't come back up—don't rush to reinstall the system. Check systemd, enable, logs, ports, and dependencies in order, and you can usually pinpoint the problem quickly.
When a VPS website connects but returns 403 Forbidden, it's usually not a network failure — the web server is refusing access. This article walks through Nginx, Apache, directory permissions, index files, SELinux, reverse proxies, and common WordPress scenarios step by step.
After installing Fail2Ban on a VPS, SSH suddenly stops connecting — possibly because your own IP was mistakenly banned by the sshd jail. Common causes include wrong passwords triggering maxretry, dynamic IPs not added to ignoreip, overly strict jail configuration, and old bans restored from the database after a restart. This article walks through troubleshooting layer by layer: console/VNC self-rescue, fail2ban-client status, unbanip, ignoreip, jail.local, dbpurgeage, and preventing another self-lockout.
After a VPS website is put behind Cloudflare, a 521 Web Server Is Down or 522 Connection Timed Out error usually does not mean Cloudflare itself is broken. More often the origin refuses connections, nothing is listening on 80/443, a firewall or security group is blocking traffic, Nginx/Apache/Caddy is not running, the origin is overloaded, origin-pull IPs are blocked, or the TLS/proxy configuration does not match. This article troubleshoots layer by layer, starting from the difference between 521 and 522, then curl to the origin, Cloudflare IPs, Nginx logs, firewalls, and timeouts.
When starting a service on a VPS you see Address already in use, EADDRINUSE, bind() failed, or Docker port is already allocated. Common causes include an old process that never exited, Nginx/Apache/Caddy grabbing 80/443, Docker port mapping conflicts, IPv4/IPv6 dual-stack listening confusion, duplicate systemd service starts, or misreading TIME_WAIT. This article troubleshoots layer by layer using ss, lsof, fuser, journalctl, Docker, and Nginx logs.
When Nginx, Node, a database or a self-hosted service on a VPS reports Too many open files, common causes include the process running out of file handles, systemd LimitNOFILE not taking effect, /etc/security/limits.conf only affecting login sessions, a mismatch in Nginx worker_connections, application connection leaks, or too many logs/sockets. This article troubleshoots layer by layer using ulimit -n, /proc/PID/limits, lsof, systemctl show and sysctl.
VPS CPU looks low, but SSH lags, the website is slow, database queries are slow, and %wa/iowait is high in top. Common causes include log explosions, random database reads and writes, Docker container logs, swap thrashing, backup compression, overly long disk queues, or host storage contention. This article locates the problem layer by layer using top, vmstat, iostat -x, iotop, pidstat, du, and journalctl.
