DNS and NTP strange problem

 
Thread Tools Search this Thread
Homework and Emergencies Emergency UNIX and Linux Support DNS and NTP strange problem
# 1  
Old 06-24-2010
DNS and NTP strange problem

Hi,

I'm having problem with my DNS and NTP.

For the DNS:
I'm using RHEL4, Bonding and heartbeat.
Bond0 IP address: 192.168.1.10
Heartbeat or Bond0:0: 192.168.1.50

Currently, I use my heartbeat ip address for my DNS or in named.conf

Everytime I reboot my server, it says that dns/named successfully started but when I type nslookup in shell, it did not show anything at all. But when I restart named by using "/etc/init.d/named restart" I able to nslookup my dns configuration.

I'm confused, what's going on? shouldnt redhat startup process starts my named? why do I need to restart again.. anyone can help?
--------------------------------------------------------------

For the NTP:
I'm using RHEL4, Bonding and heartbeat.
Bond0 IP address: 192.168.1.10
Heartbeat or Bond0:0: 192.168.1.50

I use NTP for server and client. I set my NTP configuration using my heartbeat IP (192.168.1.50)

I able to sync from server to client, the problem is also after I reboot the server. the ntpd is running, but when I trace the wireshark from my server, actually my server send the packets by using my bond0 ip address (192.168.1.10) instead using my heartbeat ip (192.168.1.50).

But when I restart ntpd "/etc/init.d/ntpd restart"

I trace again from wireshark, the server starts to send packets using my heartbeat ip address (192.168.1.50) which is the way it supposed to be.

I'm confused, whats happening here? is it because my named and ntpd starts before the heartbeat starts during the startup process when I reboot the server?

is there a way to tweak the named and ntpd to start after the heartbeat starts?

Thank You.
# 2  
Old 10-01-2010
I think that you're probably right - it might very well be a start order issue.

To change the order, since you're running RH, edit the start scripts for named and ntpd in /etc/init.d and find the line:

# chkconfig: -

The two numbers after that are the numbers that control the start, and the stop, respectively.

For example, if it's "13 87" for named, you will be able to see that there's a symlink in /etc/rc3.d called S13named (and in all the other run levels that named is configured to start in) and also one called K87named in /etc/rc6.d (and the runlevels that named is configured to be stopped at).

Edit the first number to increase the number to start named and ntpd after the bond. Increase the second number to have those processes shut down before the bond.

Then you can use "chkconfig --del named" and "chkconfig --del ntpd" to remove the existing symlinks, and then "chkconfig --add named" and "chkconfig --add ntpd" to add them with the new orders.
# 3  
Old 10-05-2010
under some distros, "servers" has to be enabled :

for instance the word ENABLE has to be in /etc/default/serverd (ntp or other services)

please check that .

also, is it to much for u to give us a ls -R of your rc?.d and the likes ...

order issue / scripts executable or not ...

regards
# 4  
Old 10-05-2010
I have had this problem with SuSE when using NetworkManager and DHCP. The Named comes up before the DHCP has successfully registered itself and named then does not work. My bodgy fix was to add a service restart for named in the local init file which initalizes last. I then stopped using NetworkManager and everything seemed to work again - I presume the static network scripts probably work better with their startup sequence ordering...
Login or Register to Ask a Question

Previous Thread | Next Thread

10 More Discussions You Might Find Interesting

1. UNIX for Beginners Questions & Answers

NTP synchronised problem in our Centos 7.6 node

Someone, please help on this issue:- Note : for security reason i didn't mention hostnames and ips. ============================================================================== # ntpstat unsynchronised polling server every 1024 s Ntpstat showing unsynchronised. ... (29 Replies)
Discussion started by: shanmugaraj
29 Replies

2. AIX

NTP time problem

I got an ntp time problem on AIX server. os version is AIX7.1 OS LEVEL 7.1.0.0 i got below output,when i run the below command bash-3.2# ntpdate -dv XXXXXXXXXXXXXXXXXXXXXXXX 4 Dec 12:50:49 ntpdate: 3.4y transmit(xxxxxxxxx) receive(xxxxxxxxx) transmit(xxxxxxxx) receive(xxxxxxxxx)... (9 Replies)
Discussion started by: murali969
9 Replies

3. HP-UX

Setting up NTP HP-UX clients from solaris NTP server

Hi I wonder if its possible to setup NTP clients running HP-UX o.s. from a solaris 10 NTP server? FR (3 Replies)
Discussion started by: fretagi
3 Replies

4. UNIX for Advanced & Expert Users

NTP Problem

Hello every one i want to setup a ntp on sco unix 5.0.6 and i don't have any time server how can i do this with local time (1 Reply)
Discussion started by: kaydream
1 Replies

5. Solaris

NTP problem.

Hi, I have edited the /etc/inet/ntp.conf file to point to a new time server. When I restart ntp on Solaris 10 using 'svcadm disable/enable ntp' it still points to the old ntp server when I run a 'ntpq -p'. Any ideas? Thanks. Sparcman (4 Replies)
Discussion started by: sparcman
4 Replies

6. Red Hat

NTP sync problem

Hi, I have 4 machines and all are NTP configured with same ntp.conf file. But 3 machines are in sync with Time server and 4th machine is not contacting & sync with the NTP server. Please see the below details. # service ntpd stop Stopping ntpd: ... (0 Replies)
Discussion started by: uday123
0 Replies

7. AIX

NTP Client problem

Hello everybody! When I run command "ntpdate" the following error occures "ksh: ntpdate: 0403-006 Execute permission denied. " # oslevel -s 5300-08-06-0918 oslevel -r 5300-08 eny suggestion friends? :-) (3 Replies)
Discussion started by: wwwzviadi
3 Replies

8. UNIX for Dummies Questions & Answers

NTP problem

All - i tried to sysn the clock from our server to get the timing and the server IP has been added in the ntp.conf file and i have some problems here. if some1 help me to solve this problem. 1. i can ping server 2. i was unable to ntpq -p (time out, no response from remote server, it looks... (2 Replies)
Discussion started by: March_2007
2 Replies

9. Solaris

NTP Authen Problem

Dear All I'm solaris user . I need to setup NTP authentication for my office server but I have some question for ntp setting up. I need to know if i set authen NTP the client that doesn't set for authen key can sync time from authen ntp server or not ?? when i setup the authen at ntp server ... (1 Reply)
Discussion started by: goasutlor
1 Replies

10. HP-UX

NTP problem

Hi, I have a problem using NTP. We have a setup as following : client --> HP Server(NTP) --> DMZ server --> DMZ server --> Internet Only some clients and our HP server is using NTP, the others are using different software. Now I get an error every day : Oct 1 13:17:15 nlunx1n1 xntpd:... (4 Replies)
Discussion started by: davidg
4 Replies
Login or Register to Ask a Question