SUMMARY: NIS+ Error Message

From: Robert Benites (benites@woodfin.cs.unca.edu)
Date: Tue Aug 04 1998 - 12:49:12 CDT


My original question is posted at the end of this message.

Thanks to all who responded:

        john@starinc.com (John Malick)
        Stuart Kendrick <sbk@fred.fhcrc.org>
        <EFernandes@talisman-energy.com>
        Bismark Espinoza <bismark@alta.jpl.nasa.gov>
        <Unixboy@aol.com>
        <EFernandes@talisman-energy.com>

There was some confusion because I did not adequately inidicate my
working environment. After assistance from John Malick
(john@starinc.com) I realized I should ahve said that the workstation
I had screwed up is a standalone NIS+ machine, it is the master root
server, and the *only* machine in the domain.

Once John Malick established that fact, he gave me a laundry list to
clean up my mess.

Here is the text of his message:

> Since you already made the changes to the /etc/hosts file and
> others, leave them. Next you need to strip off NIS+ so its not
> running and rebuild this machine as a root master. Again, you only
> need to do this if the machine is a root master and if you have no
> NIS+ clients then you don't have to worry about them.
>
> 1. The first thing I would do is to dump the NIS+ unless you have
> all the NIS+ info in the local /etc files, i.e. hosts, passwd,
> shadow, etc. If you have added NIS+ info with adminsuite or admintool
> you will have to dump the NIS+ tables with the nisaddent command. If
> you've done everything with the local files then you do not need to
> worry about this first step.
>
> 2. Clean out all NIS+ info:
>
> a. cp /etc/nsswitch.files /etc/nsswitch.conf
> b. /etc/init.d/rpc stop
> c. rm -f /etc/.rootkey
> d. rm -rf /var/nis/*
> e. rm -f /var/nis/.*
> f. /etc/init.d/rpc start
>
> 3. Rebuild NIS+ server:
>
> a. /usr/lib/nis/nisserver -v -r -d domain_name.
> (remember to end the domainname with a .
> it needs to be fully qualified, i.e.
> your.domain. <- that is not a period for
> the end of sentence but fully-qualifying.
> Answer the questions.
>
> b. Now populate the NIS+ tables:
> i. cd /etc
> ii. nispopulate -v -F
>
> 4. Reboot the server and all should be good.
>
> Again, you'll have to do some additional stuff like dump the tables if
> you've made changes to NIS+ using adminsuite or admintool.
>
> Hope this helps

And help it did. It fixed my problem. Thanks John!

-- Bob

-------------------------<Original Post Follows>-------------------------
> I have one Sun workstation running Solaris 2.5 and NIS+, most run NIS.
>
> We recently made some network changes which required changes to
> /etc/hosts and other files.
>
> I made my changes in single-user mode and when I tried going to
> multi-user I received the following message:
>
> nis_cachemgr: Erro in reading NIS Cold Start file: /var/nis/NIS_COLD_START
>
> later on I see:
>
> hosts.org_dir: NIS+ servers unreachable
> group.org_dir: NIS+ servers unreachable
> passwd.org_dir: NIS+ servers unreachable
>
> The system seems to be OK when if finally comes up, but I'm assuming I
> need to do something like regenerate the NIS+ databases, or
> /var/nis/NIS_COLD_START.
>
> I have little contact with NIS+, most other Sun's are running NIS. I'm
> not sure what to do to correct this error condition. I've tried using
> tools like nisstat, niscat, but they yield error messages that "NIS+
> servers unreachable"
>
> Any suggestions?



This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:12:45 CDT