This regards a number of CentOS 4.8 servers. They were all installed with generic options, have been in service for a long time, and haven't had any changes made in over a year, other than a change to /etc/resolv.conf, which is the reason for this post. We want to decommission an older DNS server, and are pointing everything to the replacement servers.
After changing /etc/resolv.conf, all of these systems report the proper name server when using nslookup or dig. However, whenever syslog attempts to access our logging server, it is requesting an A record lookup from the old name server. I can't blow up our syslogging, so for the time being, we are leaving the old DNS server operational.
I have done some extensive searching, but am not finding anything that resolves the problem. It is as though nscd is running, but it isn't. It may be that there is a similar caching service for syslogd, but I can't find it. Whatever it is, syslogd wants to look to our old DNS server for name resolution each time it attempts to contact the syslog server.
A ps -ef does not list named, nscd, or dnsmasq. A ps ax has the same results. Running chkconfig --list doesn't show any of those three being started.
The servers in question need to stay up and available, so I cannot reboot them or restart the network services to see if that clears the problem.
Our syslog.conf file only has one change from the default:
Any ideas why syslogd seems to be caching up the old name server information, and more importantly, how to get it to use the correct name server without either rebooting the box or restarting all the network services? To be honest, I don't even know if either of those actions would correct the problem.
Thanks in advance for any help.
Issues related to configuring your network
2 posts • Page 1 of 1