Visvanath Ratnaweera

> Initial installation went OK.
On what? Mac OS X, Win???, Solaris, BSD, Apache, IIS, ? (your one here)

> got to install.php
Now, _who_ went there (where)?

Sorry sir, we are clueless

For the usual questions here on dimensioning servers, performance and scalability we can't give specific answers simply because the "speed" is something very individual.

I would like to ask our developers whether a "benchmarking suite" for Moodle makes sense. Here are some sketchy thoughts:

Szenario A
- a client driven skript where you can configure the "load". There is some correlation between this "load" and say the number of simultanous users in a real situation.

- to simulate "very heavy" loads 'fcourse one can add more clients.

- if it to simulate Moodle at work, the initial content of the installation must be defined

Szenatio B
- we use the standard benchmarks for the underlying components, like: web-server, database and PHP.

- through experience find out which set of figures are sufficient to service a certain number of simultanous users.

What do you think? Is anything like this is already available? Or is it just a crazy idea?

Do you say, you can use phpMyAdmin to manage the database, but Moodle fails to connect to the database?

That is quite unusual. Both are just PHP applications.

> its a pain on the ass running P4 with single processor 2gb of ram. PHP eats up the CPU till it dies, even running on accelerators. By the way we had 4,000 students. 150 students log at the same time kills the dedicated server.

Doesn't sound very encouraging, specially for places where the money _is_ an issue

Could you tell us more about the P4/2GB setup, operating system, database, Moodle version, etc. ?

I wonder whether a scheme of benchmarking is available. Will start a new thread on that.

So you think that there are technical methods of making sure that nobody breaks the law "inadvertently"?

A side question: If you are going to "protect" teachers from this pitfall, why not the administrators?