On Tue, Mar 01, 2005 at 04:12:57PM +0100, Marco Corvo wrote:
> In the view you propose, I should make a list of requirements, send it
> to all site managers whose site I would like to "exploit", wait for the
> "green light" and then start installing. Did I understand correctly?
>
> Then, maybe, the specific rpms conflict with the standard configuration...
No this is not what i suggested at all.
I was only talking about the *system* software since someone asked what can
you expect to see in a node, ie does it have md5sum/ldapsearch/ssh/ksh/whatever?
It is about what the "standard" configuration is......
I as a sysadmin have a list of what to install in every new installation
in my kickstart file or whatever else i might be using.
This list is the union of what I think is needed and what the local
users have asked for.
This list does not include everything that every VO wants/needs, because
I never thought that anyone will need the ogg libs in a cluster
node for example.
If your VO wants me to have some *system* software installed it then it needs
to somehow communicate that need to me and since as most sysadmins i am
lazy i would prefer that to be automated somehow :)
One way (for me) to automate it is by having a list in a "nice" format
an rpm or even an xml document[1] that i can use to install the base system.
Do not expect the sysadmins to install everything from a RHEL/SL/debian/whatever
some distributions (debian for example) have so many packages that it's
unlikely that it will happen.
Cheers,
Kostas
[1] the yum metadata format springs to mind here, which is portable enough to
get an rpm/dpkg/whatever with some simple scripts.
|