Green server build

[Robert Olson]

Very cool Ti - thanks.

It sounds like you'd recommend that we install things like stable snapshots 
of the new VV code in /disks/space0?

At 11:54 AM 8/27/2002 -0500, Ti Leggett wrote:
>Alrighty, I *think* the green server build is a go. If you want to login
>to one of the green servers, please try vv2.mcs.anl.gov and hammer away
>at it. And please let me know if there are any problems or if you really
>dig it.
>
>Here's what it has:
>
>- Globus-2.0 installed in /my/packages/globus (set GLOBUS_LOCATION
>there)
>- GPT-1.0 installed in /my/packages/gpt (set GPT_LOCATION there)
>- python1.5 and python2.2
>- wxGTK and wxPython
>- pyGlobus-0.9.6 installed in /my/packages/pyglobus, but there's a link
>to it from /usr/lib/python2.2/site-packages/pyGlobus so you shouldn't
>have to do anything special to use it
>- apache with mod_perl, mod_python, mod_ssl, mod_dav, and php. (this
>still needs to be configured but it should be working and running)
>- MySQL with perl, python, and php bindings (if you want a database let
>me know and I'll create it)
>- PostgreSQL with perl, python, and php bindings (if you want a database
>created let me know)
>- sudo (with a *very* limited set of people authorized for it)
>- Group writeable local disk at /disks/space0 (this is where you should
>put files that you want anyone to be able to use)
>- There is no NFS (Yeah!)
>- Home filesystems are mounted via samba from glacier at login time, so
>that means your environment will look the same as on green machines,
>*BUT* no other users home dirs are available unless they are logged.
>This is stronger incentive to use the group writeable storage space.
>- I'm trying to make these machines single purpose servers or at the
>very most serving several coupled or closely related services. Hopefully
>this will aid in finding projects and code a little easier. i.e., If
>you're looking for VV2 code a good place to look would be
>vv2:/disks/space0
>- ADSM backup is on the way but not there yet.
>- There's possibly other things that might be installed as well that I
>forgot, a good thing to do might be to do a "rpm -qa | sort | less" to
>see what all is installed. My aim is install packages either via RPM or
>in /my/packages so that it makes finding things much easier.
>
>One other note on the databases. The databases on each machine are meant
>for local development testing (doesn't mean it has to be only for that).
>If you need a production database, I'd suggest asking systems for one on
>yeah.mcs.anl.gov.