Re: diskless systems with RHEL 3
Daniel J Walsh <[email protected]>
| Newsgroups | gmane.linux.redhat.taroon |
|---|---|
| Message-ID | <[email protected]> |
It was not intended to be used that way. The code is going to look for a /root directory below the specified shared directory. So it when you specify / as the top of the shared directories it will look for /root and /snapshot as being the shared directories. You would have to change the scripts to get around this problem. I didn't just share the server OS because of the conflicts of the way the two server os and client os would work. For example, the server needs to run several services that the client doesn't (nfs, tftp). Files like /etc/exports would have to be different. Sharing of the /etc/passwd and /etc/shadow of the server is risky. I would recommend that you keep them different. Dan Alois Treindl wrote: >On Mon, 29 Sep 2003, Daniel J Walsh wrote: > > >>The easiest way to keep the clients up2date is to chroot on the server >>and run up2date. >> >> >> > >Thanks for the swift reply. It is great to talk to the developer himself. > >The redhat-config-netboot scruipt advises me to >install a complete sample client OS and then to copy it to the >diskless server, from which the diskless clients will boot it. > >I intend the sample client and the diskless server to be the same machine. > >is it really necessary to duplicate the OS for that purpose, or can I do >something like this on the server: > >mkdir /diskless >cd /diskless >ln -s / root > >and proceed from there? > >or is there another way to avoid duplication of the OS in the case when >diskless server and 'sample client' are identical? > > >-- >Taroon-beta-list mailing list >[email protected] >http://www.redhat.com/mailman/listinfo/taroon-beta-list > >