I am pretty new to Linux so please be patient with me

I have been designated to "audit" a linux cluster. The company I am contracting for has been having issues pretty often:

Their customer goes to click on a document and it says that its not available - not just one document but many. Then after about 5 minutes of down time it will be back up again.

I do not have physical access to the servers as they are in another state, I also do not have root access because everything is managed by another company and they dont know what the problem is either. I just have user access through putty.

What I am asking here is a starting point. I need to document info on 21 of these servers and they run rhel 5.2 - 5.5.

I do have info on build dates, install dates, package info, yum list installed and that is a lot of info so far.

Any suggestions would help, I know its not much for anyone to go on, but I wasnt given much to go on either, no specific error msgs or anything and of course I cant view the log files or config files but should have those later today hopefully.

I can run whatever commands you would like me to in order to provide more information please just let me know what commands you would like me to run. I do have tons of info on each server already.

Thank you!

I would say the starting point is logs..
monitoring some log files whilst trying to generate log traffic by accessing the documents that will fail. (is it the same document/s)
ck on a document and it says tha
How are the document's shared? Are the servers under big loads when they are experiencing problems?

Ultimately, even if you DO manage to diagnose a problem without root access, you aren't going to be able to fix it..

I figured that the logs would be a starting point, the only problem is that I cant even view the logs without root access...

The servers are not under any extremely huge loads when this happens and the real issue is that it happens randomly (nothing is random but it looks random at this point).

What do you mean how are the documents shared? Sorry im a noob

My end goal is not to actually fix the problem but to diagnose what the problem could be, have the group that manages these servers/clusters repair it and see if that was the problem or not.

Again, any help or command suggestions to run would be great!

Thank you for your response!