We help IT Professionals succeed at work.

We've partnered with Certified Experts, Carl Webster and Richard Faulkner, to bring you two Citrix podcasts. Learn about 2020 trends and get answers to your biggest Citrix questions!Listen Now

x

Major Solaris NIS Problems

saintjoes
saintjoes asked
on
Medium Priority
444 Views
Last Modified: 2013-12-16
We have a bunch of Solaris 2.5.1 servers that
are experiencing major NIS problems.  We have installed
NSKIT 1.2 (with appropriate patches).  We have all OS patches installed.  We are having problems in these areas:
- Autmount is not working properly (same subnet, fast machines) - it timesout mounting a share
- Logging in takes a long time to authenticate (Pc and UNIX side) and sometime authenticates correct passwords incorrectly.
- PC users cannot use NIS to do hostname resolution using a winsock application.
- Ypbind loses its binding to the domain often.
Any ideas?
Comment
Watch Question

pxh

Commented:
I have a very similar config as you (a NIS master, running 2.5.1 and having the NISKIT installed). I do not have your problems. Let me know more details about your config, the above is too common.

Only answer now: for hostname resolution you usually use a DNS service. You can do it within your NIS domain also, but I have not herad that there is NIS for PC's...

Peter (pxh@mpe-garching.mpg.de)

Author

Commented:
We have NIS setup to DNS forwarding.  The client PC's are
using NIS as thier host name resolution.  The only odd thing in our setup is a SunOS master (these 2.5.1 machines are actually slaves, but are being used as NIS servers).
We have duplicated this on *3* Solaris 2.5.1 machines!



Commented:
Again, there is no obvious problem. We also had here a SunOS
NIS master running successfully for quite a while. Now I have one master + one slave under Solaris 2.5(.1) + mixed (Solaris + SunOS) clients.

Hmm, it is difficult do help you, things have to be verified step by step. I hope I am not bothering you with trivialities, but just to make shure: did you activate the right /etc/nsswitch.conf file? I.e. did you copy nsswitch.nis to nsswitch.conf an all systems?

So some more questions: do you have automatic or static NIS server recognition? I.e. does ypbind run with -broadcast options? How did you install the clients? Did you run ypinit -c and did you then specify all master and slaves? I found that methode more stable then running dynamic binding (ypbind -broadcast).

Then you have to find for a given client (a slave might be regarded as a client) which is his master, use the ypwhich command on that machine. Try to isolate the problem, e.g. shut down all slaves and run with one master only.

To the slaves: did you add the crontab entries? Here an example from mine:

#
# run the ypxfr's to syncronize aldebaran (slave) with orion (master)
#
0 * * * * /usr/lib/netsvc/yp/ypxfr_1perhour
0 1 * * * /usr/lib/netsvc/yp/ypxfr_1perday
0 2,14 * * * /usr/lib/netsvc/yp/ypxfr_2perday

You might have to edit the ypxfr scripts to support the tables you need.

Then your automount problem: if you enter "ypcat auto.direct" do you get the correct listing? I had to add en entry to /var/yp/Makefile for the support of the auto.direct file. Also check in /etc/nsswitch.conf for the line "automount:  nis files" to be present.

Then the name resolution prob: look into /var/yp/Makefile (of your master) and make shure the the line "B=-b" is active. Does the name resolution work at all, i.e. does nslookup <address> work?

Last not least are the right daemons running? On the master (Solaris):
/usr/lib/netsvc/yp/ypserv -d
/usr/lib/netsvc/yp/ypbind
/usr/lib/netsvc/yp/ypxfrd
/usr/lib/netsvc/yp/rpc.yppasswdd -D

on the slave(s):
/usr/lib/netsvc/yp/ypserv -d
/usr/lib/netsvc/yp/ypbind

and on the clients:
/usr/lib/netsvc/yp/ypbind

Hope you get any idea from that,

Peter (pxh@mpe-garching.mpg.de)

Not the solution you were looking for? Getting a personalized solution is easy.

Ask the Experts
pxh

Commented:
I am confused...
you wrote in the comment to my answer:

"Another note: Clients wore fine if they use the SunOS machine as their nis master." and "But all clients are using the solaris machine as their master."

I conclude that you switched between master and slave server and found out that only the slave server give problems.

Question: does NIS work properly on the slave(s) itselve? Is there a difference between those slave(s)?

Again: I am not nervous about your master being a SunOS system. I had this situation here too for several years (also with previous Solaris versions).

What did you precisly do to "We have duplicated this on *3* Solaris 2.5.1 machines!"? "this" meaning the slave service.

How does your /etc/auto_master look like?


Peter (pxh@mpe-garching.mpg.de)

Access more of Experts Exchange with a free account
Thanks for using Experts Exchange.

Create a free account to continue.

Limited access with a free account allows you to:

  • View three pieces of content (articles, solutions, posts, and videos)
  • Ask the experts questions (counted toward content limit)
  • Customize your dashboard and profile

*This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

OR

Please enter a first name

Please enter a last name

8+ characters (letters, numbers, and a symbol)

By clicking, you agree to the Terms of Use and Privacy Policy.