My original question was based on the following, but I have clarified our
specific problems at the end of this message:
I'm a bit confused about the relationship between an NIS master
server and an NIS slave server. Our setup is as follows:
=============================================================================
| | | | |
FileServer 1 FileServer 2 FileServer 3 FileServer 4 |
(master) (slave) (slave) (slave) |
| | | |
| | | |
clients clients clients remote clients
....if the master server goes down, ......
Question: Which one (if any) is now the NIS master?
They all seem to compete for this role......, 'ypwhich' can produce
four different values on each one, and confusion reigns.
Thank you to: George D M Ross, Edinburgh
Andie Ness, Edinburgh
Dick St. Peters, New York
David Stewart, Edinburgh
Ron Hall, Montreal
Jay Lessert, USA?
kevin@saber.com
and Bob Cunningham,
who all told me, that basically nothing is wrong at all. The
'ypwhich' command does not give the master server, but the current NIS server,
and this will change from time to time as the load on the servers goes up and
down.
Only if the master is down for a long time need I worry about defining
a new master, and transferring the database file to it from backup.
However! I should have mentioned the sort of problems which we are
encountring - sorry; here they are:
1. I created a new group in the masters /etc/group file and placed
some users in it, eg
osi_team:*:220:user1,user2.....
I went into /var/yp and typed 'make group'. It said 'updated group,
pushed group'; so far so good.
But, for some reason, the users find (by typing 'groups') that they
are only in osi_team when logged on to the master server NOT when on the salves,
or any of the clients. All the machines are running ypbind, ypinit, etc. Why
are things not getting propogated to the slaves?
2. A dot-matrix printer attached to one of the slave servers queued
jobs correctly, but would not print anything. Various errors were reported
from 'lpq' such as 'waiting for lock on dev/ttyb', 'Warning no daemon present',
' (blank status line) ', and the first job refused to become 'active'.
I spent ages with lpc> killing daemons, clearing queues and deleting
lock files manualy all to no avail. What *appeared* to solve the problem, was
doing a 'ypbind -ypset', 'ypset <master server>' on each of the slaves.
Could there be a connection with the first problem. It's very strange.
Please reply direct to:
and I will post a summary.
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:06:10 CDT