Unix (and Unix-like) operating systems use the UIC to identify users. This
is often to only means available for authentication. It is necessary for
system administrators of such systems to coordinate the allocation of UICs
and UIDs accross an organization. NFS and other network services use the
UIC to assign file access rights to users.

In the absence of coordination of UIC allocation, we could end up with a
bunch of workstations whose main user accounts have a UIC of 100. This
means that all those users can see any network visible file owned by all
the other users with the same UIC.

The only real solution is to change the UICs so no two users have the
same one. This is clearly a difficult task since you much change the
ownership of all existing files. Don't even think of backups that were
made before this.

On Wed, 6 Nov 2002, John E. Malmberg wrote:
> > [EMAIL PROTECTED] wrote:
>
> <snip>
>
> > In this example, the value 9044133 corresponds to UIC = [212,245].
> > The problem is that the sysuaf database does not effectively have an
> > identifier with a value of [212,245] and a name egal to the username.
> > This was not a problem with PathWorks but how can I deal with that in
> > Samba ?
>
> Samba on OpenVMS requires that the UIC match a username, and that the
> username have a default directory that it has write access to.

-- 
Claude Marinier, Information Technology
Defence Research & Development Canada (Ottawa)
[EMAIL PROTECTED]
http://www.ottawa.drdc-rddc.gc.ca

Reply via email to