I am new to Linux and would appreciate suggestions. I did do some sysadmin on Solaris along with Oracle dba for about 12 years, but it was a small shop so no setup like this.

We want to set up a set of servers as follows: RHEL 5, hosted environment, so part of the work is done for us, initially application test environment, we will begin in about 3 weeks.

Firewall into a web server. This first server routes user activity to backend servers depending on the desired function to perform.

Function 1.a Upload files for parsing/cleaning data/move data into a MySQL database. Nothing major here. User goes to a db server connected to a fc san for the MySQL db.

Function 1.b Flat files created from data in the MySQL database for users interested in function 2 below. Flat files will be moved moved over to a sata san not directly connected. The flat files will be read only.

When in use, not sure the best way to move files with modified content same name into production. Move the file with a different name and rename when the old file is not open?

Function 2. [This is where the hosting service says we may have problems with what we want to do.]
If the user is interested in the data in the flat files, we want to route the user to 1 of several (in the test environment there will be 2) servers - we would like some form of load balancing - apache??

We would like each server to access the same set of flat files on the sata san. The hosting service says they don't support active active clustering, and they also say trying to access the same memory will lead to troubles. Any suggestions on set up for this use?

Also on moving new and modified files into the production area as stated in 1.b - we anticipate in production concurrent users, though not necessarily in the same files that is a possibility.

We expect the database to grow to 300GB or so, and the flat file space to be perhaps 700 GB because of de-normalization of the data from the database to speed the queries. At this point, we are not sure though - some estimates show the filespace could grow into the 1-2TB range.

Finally, I am looking for good reference material. I was looking at
"Red Hat Enterprise Linux 5: Administration Security Desktop" (Richard L Peterson)
Also something on Apache or Apache Tomcat if we want to use Apache or Tomcat.
Any suggestions?

Thanks

This is pretty comprehensive (as it should be )
http://www.linuxtopia.org/online_boo...ion/index.html

Thanks Chris, as you said, it looks like it is comprehensive. I'll go through it.