I was wondering if anyone has ever come across this before, as I'm pretty stumped to be honest, and my server admin knowledge isn't particular good so I'm not sure what could even be wrong, let alone how to fix it.

Basically,

Thursday last week everything was fine on our server. I come in on Friday and it's a mess: php extensions are missing/not working, apache modules are gone. (e.g. oci_* was gone completely, odbc_ not working but still there, the apache ntlm_auth for single sign on was gone and so the website wasn't even loading in IE).

I'm ruling out anything deliberate because it's just incredibly unlikely.

The only thing that really happened between thursday & friday is that on thursday evening one of the network guys did a RAM upgrade on the server and restarted it. That's it, nothing else.

Now I'm wondering if somehow those extensions and such which we installed months ago were somehow only saved in a local memory of sorts, and a restart has wiped them? But we installed them all as root, so I don't see why it should be any different from installing anything else.

It makes little/no sense to me.

To expand on an example of something that's gone very wrong, the php odbc_ extension:

It's still on the server, it doesn't return undefined function or anything. But it just cannot connect to the datasource any more. I've tested it through the command line and it's working perfectly fine with that datasource and login details, but all of a sudden having it in the php odbc_connect() function and it just can't connect. ( [S1000][unixODBC][FreeTDS][SQL Server]Unable to connect to data source. )

But unixODBC is set up fine. Like I say i've tested it all through the terminal and it can connect, and we've not changed anything, it's just now all of a sudden not working through the PHP function.

Anyone have any ideas whatsoever as to what could be going on?

This is on CentOS 5.x by the way.