Linux Server Admin Needed — Fix Recurring Website Downtime on HestiaCP

We manage a web hosting server running

160+ WordPress websites that keeps experiencing a recurring issue: all websites go grey/blank every few hours after a server restart.

A manual restart temporarily fixes the problem, but it returns within 1-4 hours.

We need an experienced Linux server administrator to diagnose the root cause and implement a permanent fix.

## Server Details

• Hosting: Linode (dedicated server)
• OS: Ubuntu 22.04 LTS
• Control Panel: HestiaCP (latest version)
• Web Stack: Nginx (reverse proxy) → Apache 2.4 → PHP-FPM 8.2
• Database: MariaDB 10.11
• RAM: 32 GB
• Websites: ~160 WordPress sites
• PHP-FPM Pools: 109 (ondemand mode, 8 max_children per pool)

## The Problem

• All websites simultaneously go grey/blank (empty HTML response) within 14 hours of a restart
• Restarting the server or Apache + PHP-FPM temporarily fixes it
• All services (Nginx, Apache, PHP-FPM, MariaDB) show as "active (running)" even when sites are down
• No errors in WordPress debug logs
• Nginx error log only shows harmless ssl_stapling warnings

## What We've Found So Far

• Apache error log shows repeated `"Primary script unknown"` (proxy_fcgi) and `"Broken pipe: Error dispatching request to : (sending empty stdin)"` errors when sites go down
• Apache returns HTTP 301 redirect loop (WordPress redirecting HTTP → HTTPS) when tested locally via curl during downtime
• Nginx proxy config (``) has

no `proxy_set_header` directives — no `Host`, `X-RealIP`, `X-ForwardedFor`, or `X-ForwardedProto` headers are being passed from Nginx to Apache

• PHP/WordPress loads fine via CLI (`php r "require 'wp';"` returns OK)
• Apache `MaxRequestWorkers` was set to 150 for 160 sites (recently increased to 400 — monitoring if this resolves it)
• `KeepAlive` was Off (recently changed to On)
• PHP-FPM socket files exist and are accessible

## What We Need

1.

Root cause diagnosis — determine exactly why sites go down after a few hours

2.

Permanent fix — implement a solution that prevents the recurring downtime

3.

Nginx proxy headers — safely add `proxy_set_header` directives (Host, X-Real-IP, X-Forwarded-For, X-Forwarded-Proto) to HestiaCP Nginx templates without breaking sites (our first attempt caused 500 errors)

4.

Apache/PHP-FPM tuning — optimize configuration for 160 WordPress sites on 32GB RAM

5.

Documentation — brief summary of what was changed and why

## Ideal Candidate

• Strong experience with HestiaCP administration
• Deep knowledge of Nginx reverse proxy + Apache + PHP-FPM stack
• Experience managing 100+ WordPress sites on a single server
• Familiarity with Linode hosting
• Available to work on this urgently

## To Apply

Please describe your experience with HestiaCP and similar Nginx/Apache proxy issues. Bonus if you've dealt with "Primary script unknown" errors before.

Contract duration of less than 1 month. with 30 hours per week.

Mandatory skills:

Linux System Administration, System Administration, Linux, Bash, NGINX, Business with 10-99 Employees