SUMMARY: Wierd DNS / NIS problem

From: Nils K. Schoyen (
Date: Tue Jun 01 1999 - 01:51:06 CDT

Hello again!

Thanks to Roy Culley and Gustavo Chaves who responded to this mail.
Unfortunaltely, their suggestions did not solve our problems :-(

The problem was solved like this:

  o Power down ALL NIS-client
  o Power down NIS-server
  o Power up NIS-server
  o Power up the NIS-clients

It seems as if the NIS-clients where unable to succesfully re-establish
the connection to the NIS-server when the server had been down for more
than e.g. two hours. Taking the NIS-server down for e.g. 30 minutes or so
works fine, but when the server is down for a longer period, the NIS-clients
fails to re-establish the connection.


Nils K. Sch°yen

Original posting:
> We are having some wierd problems with DNS...
> DNS server: Solaris 7. Ultra 1/170.
> NIS server: Solaris 7. Ultra Enterprise 250.
> The NIS server use the domain name resolver for hosts
> not in the current domain (B=-b in /var/yp/Makefile).
> The NIS-server, NIS-clients and all other hosts uses
> /etc/resolv.conf. The first nameserver entry points to
> our DNS server, second entry point to slave server.
> Up until yesterday everything was OK. But now, strange
> things happen...
> DNS-server-Sol7# ping
> is alive
> NIS-server-Sol7# ping
> ping: unknown host
> NIS-client-Sol7# ping
> ping: unknown host
> PC c:\> ping
> is alive
> The /etc/nsswitch.conf entry for all hosts is
> hosts: nis [NOTFOUND=return] files
> If I put dns in after nis, everything works fine...
> We tried to put in lines in named.boot to only allow zone transfers to
> the master and slave name server two days ago. These lines has now been
> removed again due to the strange behaviour of DNS.

This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:13:20 CDT