Re: [Setup-tool-hackers] Re: NIS Hacking of users-conf



Arturo Espinosa Aldama wrote:

> Well, the idea is that, if the backend detects that it is running under a
> NIS server, then a nis flag in the XML representations would be raised for
> all NIS-collected information (only users and groups). We would have to
> see how to make a difference between NIS users and system users in the
> interface.
This is achieved by a call to ypdomainname.  If it has a domain name,
then you can pretty much assume the workstation is running under NIS,
and act accordingly.  I am still getting my head round the perl, and how
to integrate this (although I made a fair amount of hacking progress
last night).

> If it is a NIS client, use ypcat to read the NIS user and group databases.
> Then, see which commands are available to make remote modifications, such
> as yppasswd, and decide then which parts the interface and backend will
> allow to be modified.
Indeed.  If it is running on a workstation, I suspect this will be
fairly minimal, however, if the tools (which is part of the idea of
OpenConnect) are running on the NIS server, then it should be able to
modify and re-make the NIS maps (and push out to any other NIS
servers).  

> Regarding the Solaris and IRIX ports, be our guests. I hope there come no
> version-specific issues with these two flavours of UNIX.
The backend should be fairly simple (although I only have IRIX NIS
clients, not an IRIX NIS server atm).  I can certainly try it on our
development NIS server, which has 30K Users &) 

I would like to see the ability of bulk creations and deletions built
into the UI, and quota, things I have just spent a while doing ;)  I
think this fits in more with the Openconnect model of being able to use
the HST to manipulate larger networks though.

Cheers

Andrew

-- 
Andrew Rawlins - University of Southampton Computing Services
Research Systems Support and Development 
023 8059 5546



[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]