Performance Glossary
Apache, and Where Its Concurrency Ceiling Sits
You need to explain why a server that is comfortable at ten concurrent visitors falls over at two hundred, and what changing it would actually buy.
The short answer
Apache is the long-running open-source web server that assigns each connection to a worker process or thread, which is what makes its per-directory configuration possible and what puts a hard ceiling on how many visitors it can hold at once.
That ceiling is arithmetic, not mystery: available memory divided by the resident size of one worker. Everything past that number queues, and queueing is what your visitors experience as the site going down.
By the HostingFast team · Reviewed 15 August 2026
100+
Terms, measured not asserted
2 min
To read one entry
Plain
Commands you can run
24/7
Engineers on shift
Modules plus configuration you can set folder by folder made it the shared hosting standard for two decades, and that legacy is why almost every tutorial you will ever read is written in Apache's vocabulary — rewrite rules, virtual hosts, .htaccess.
The cost shows up under concurrency. A process-per-request model holds memory for the whole duration of a slow client, which is why event-driven designs pulled ahead on exactly the workloads that grew fastest. Our own choice, LiteSpeed, keeps the Apache configuration surface and puts an event-driven engine underneath it.
Working out the ceiling on paper
Take the resident memory of one worker, multiply by MaxRequestWorkers, and compare that against the RAM the machine actually has. If the product is larger, the server will start swapping under load long before it reaches its configured limit, and swapping turns a slow page into a stalled one.
Under prefork with mod_php the worker carries the PHP interpreter and every loaded extension, which is why the same box holds far fewer concurrent WordPress requests than it does concurrent image requests. Serving static assets from a different layer is therefore a capacity change, not a cosmetic one.
Reading server-status while it happens
mod_status gives you a live scoreboard: how many workers are idle, how many are reading a request, how many are writing a reply, and how many are stuck in keep-alive. When the idle count sits at zero during your peak hour, you have found the ceiling without needing to theorise about it.
Watch the W and K columns in particular. A queue full of workers writing to slow mobile connections is a different problem from a queue full of workers waiting on PHP, and the fix for each is in a different part of the stack.
What Apache's ecosystem is genuinely worth
Two decades of recipes assume Apache semantics. Rewrite rules from a 2011 migration guide still run unchanged on our LiteSpeed servers, because LiteSpeed reads the same .htaccess directives. That compatibility is a real asset and it is the reason we did not simply move everything to a server that discards it.
It also means the debugging vocabulary transfers. Error messages, module names and directive syntax you already know keep meaning the same thing, which shortens the distance between noticing a problem and knowing where to look.
Proving the swap was worth it
Load-test the same URL at the same concurrency before and after any server change — a fixed number of connections, a fixed duration, the same origin. What you want to see is the point where response time starts climbing move to the right, not simply a lower figure at one visitor.
Then confirm nothing broke: request a handful of URLs that depend on rewrite rules and read the status codes. A configuration migration that reports a better benchmark and quietly 404s your permalinks has not improved anything.
Server internals, explained at ticket length
These entries assume you can read a config file and want to know why the default is what it is, not merely which button to press in a panel.
Every plan runs LiteSpeed with the cache compiled in rather than a plugin bolted on, on NVMe storage that the entry tier shares with the top one.
- Capacity worked out on paper
- The live scoreboard to read
- Compatibility claims made specific
- Load-test method described
Why HostingFast
Standard on every plan
Capacity as arithmetic
The ceiling on concurrency written as a calculation you can do with two numbers you already have.
Live evidence, not folklore
Where a server exposes its own state, the entry names the module and the column worth watching.
Compatibility stated precisely
Which Apache behaviour survives on our platform, so old recipes can be trusted or discarded on evidence.
Different queues, different fixes
Workers stuck writing to slow clients and workers stuck waiting on PHP are separate problems with separate answers.
A benchmark method that holds
How to compare two server configurations without accidentally measuring your own network.
Pointed at the next layer
Each entry names the term that explains whatever your reading has just turned into a question.
Quick Start
From order to online
- 1
Measure one worker's resident size
Read it from top or ps during your busiest hour, not at idle. The idle figure flatters the calculation by a wide margin.
- 2
Multiply out against real RAM
Worker size times the configured worker limit tells you whether the ceiling is memory or configuration. One of the two is always lower.
- 3
Watch the scoreboard at peak
Open server-status during your busiest hour and look for an idle count of zero. That single reading settles most arguments about capacity.
Built In
Loaded onto every plan
- LiteSpeed compiled into the server, not a caching plugin bolted on afterwards
- NVMe on every tier — the entry plan runs the same drives as the top one
- PHP version selectable per site, switched from the control panel in seconds
- SSH, Git and Composer on the developer-focused plans
- DDoS filtering absorbed at the network edge, before it reaches your workers
- 99.9% uptime as the target, monitored around the clock
- Staging you can clone, break and throw away before anything reaches live
- cPanel, so every script and tutorial you already have still applies
- A human on support at any hour, including for the awkward questions
- Zero setup fees, on any plan, ever
Frequently Asked
What people ask us most often
Apache or nginx — which should I put in front of a PHP application?
They optimise for different things: Apache for per-directory flexibility and its .htaccess ecosystem, nginx for holding thousands of connections cheaply. For most site owners the honest answer is whichever your host runs competently. Ours is LiteSpeed, which takes the Apache configuration surface and the event-driven model together.
Will I ever touch Apache configuration directly on shared hosting?
Only through .htaccess and the panel, because the server itself is maintained for you. That surface is deliberately thin and genuinely powerful — rewrite rules, headers, caching directives and access control all reach it, which covers nearly everything a shared hosting customer needs to change.
Can I run a restore without opening a ticket?
Yes. Every plan takes a daily backup and the restore runs from your own panel — files, databases or both — at whatever hour you discover the problem. Keeping an additional copy somewhere off our network remains good practice, and nothing in the platform stops you doing it.
How many sites can share a single account?
From the Turbo tier upwards you can run several, each with its own domain, mailboxes and certificate inside one account. If the extra sites belong to clients rather than to you, reseller hosting is the better shape: it keeps each account properly walled off from the others.
Keep reading
Nginx
Event-driven serving and reverse proxying, and why it never reads .htaccess.
LiteSpeed
The Apache-compatible server with the cache built in, and how to confirm it is hitting.
How to Test Your Website Speed
Which tool answers which question, and how to take a baseline you can defend.
WordPress Hosting
WordPress with LiteSpeed caching, staging and updates handled for you.
VPS Hosting
KVM virtual servers with root access, DDoS filtering and one flat monthly price.
Changing hosts? Run through our checklist first.
A straightforward sequence for a switch your visitors never feel: which files move first, how to shift email across without losing a single message, the right moment to repoint DNS, and the two mistakes behind almost all the downtime we get asked to rescue.
Keep the Apache rules. Lose the ceiling.
LiteSpeed reads your existing .htaccess and serves cached pages without waking PHP, on NVMe across every tier.
View WordPress Hosting plans