Take the 2-minute tour ×
Server Fault is a question and answer site for system and network administrators. It's 100% free, no registration required.

I'm curently hosting my clients web-sites on some vps servers. What I would like to know, how many visitors can use the multiple web sites concurrently, before I need to migrate each of the sites to their seperate servers?

So in laymans terms, how much RAM does a web-page user take on average? For example, how many concurrent visitors can I have on my server, running an Apache server, with 512mb RAM?

share|improve this question

2 Answers 2

up vote 2 down vote accepted

There is no way to answer your question. It completely depends on how efficient the code that generates the page is, how user visits are spread out over time, and how the setup is tuned.

You must benchmark your servers yourself. The 'best' way is to run a comprehensive test from PC's you control, using tools like ab. A simpler, easier way to get a rough idea is to use a hosted cloud stresstester like Loadimpact.com.

share|improve this answer
    
+1: simulating expected use patterns is the key. Rather than pushing a live VM to its limit which could cause you down-time though, do your initial tests on a VM on a local machine (your own desktop if it has enough RAM) - this won't be entirely accurate (your local I/O subsystem with be different and under much different load patterns to the VM hosts' IO arrangement, amongst other differences) but it will give you a fair idea and a place to start refining it from. –  David Spillett Dec 26 '10 at 13:42

It depends on the scripts used to generate your pages, if they do image manipulation for example they will consume more memory, if you just do some db queries and display the results then it will not consume to much and if you will server static html pages then your memory consumption will be very low.

It also depends on how much memory will be available to apache, if you have your database server and other services on the same server then less memory will be available to apache.

And most of the times the bottleneck is the database not apache, if you want a roughly estimate your server will still probably serve around ~30 or more request per second.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.