Hi all,
As far as I know the secondary DNS servers in the IP config of a Windows OS will not be used unless the first is not available right?
Thing is, I've got a testlab here with a S2k3 AD/DNS/DHCP/RRAS(nat) server. DNS has been configured with a forwarder, being the ISP's DNS server. The server is a virtual machine. The testlab also contains one Vista client, it gets its IP config from DHCP on the S2k3 server (VM), including DNS settings.
The testlab functioned perfectly like this for a few weeks, with the S2k3 VM being the only DNS server in Vista's IP config. No resolving problems whatsoever.
One week ago I made the following change:
As the VM sometimes needs to be offline for whatever reason, I configured DHCP with a long leasetime. I also configured the host OS with ipmasq and dnsmasq (NAT and caching/forewarding dns in linux), and added a secondary gateway and DNS server to the DHCP options. This way the Vista client would still be able to browse the Internet even if the S2k3 server (VM) is not available. Tested this new setup by suspending the S2k3 VM; worked like a charm, everything was being resolved just fine except the hosts in the VM's DNS table, just as expected. Tracert showed the linux host was being used as a gateway. After resuming the VM, internal hosts were resolvable again, and it was also used as a gateway again.
Then, one day later, the VM's hostname suddenly wouldn't resolve anymore from the Vista client. I started troubleshooting, but all over sudden everything seemed to work again. At first I thought it was a glitch, but from time to time, this strange behaviour kept occurring. So I enabled logging in the dnsmasq daemon in linux, and sure enough, it was being used to resolve DNS queries from the Vista client from time to time, even though the primary DNS seems fine.
I'll set up a script on the linux host to check DNS on the VM every five minutes using digg. This way I'll prove the use of the secondary dns while the primary is available.
If anyone can shed some light on this, please do.
As far as I know the secondary DNS servers in the IP config of a Windows OS will not be used unless the first is not available right?
Thing is, I've got a testlab here with a S2k3 AD/DNS/DHCP/RRAS(nat) server. DNS has been configured with a forwarder, being the ISP's DNS server. The server is a virtual machine. The testlab also contains one Vista client, it gets its IP config from DHCP on the S2k3 server (VM), including DNS settings.
The testlab functioned perfectly like this for a few weeks, with the S2k3 VM being the only DNS server in Vista's IP config. No resolving problems whatsoever.
One week ago I made the following change:
As the VM sometimes needs to be offline for whatever reason, I configured DHCP with a long leasetime. I also configured the host OS with ipmasq and dnsmasq (NAT and caching/forewarding dns in linux), and added a secondary gateway and DNS server to the DHCP options. This way the Vista client would still be able to browse the Internet even if the S2k3 server (VM) is not available. Tested this new setup by suspending the S2k3 VM; worked like a charm, everything was being resolved just fine except the hosts in the VM's DNS table, just as expected. Tracert showed the linux host was being used as a gateway. After resuming the VM, internal hosts were resolvable again, and it was also used as a gateway again.
Then, one day later, the VM's hostname suddenly wouldn't resolve anymore from the Vista client. I started troubleshooting, but all over sudden everything seemed to work again. At first I thought it was a glitch, but from time to time, this strange behaviour kept occurring. So I enabled logging in the dnsmasq daemon in linux, and sure enough, it was being used to resolve DNS queries from the Vista client from time to time, even though the primary DNS seems fine.
I'll set up a script on the linux host to check DNS on the VM every five minutes using digg. This way I'll prove the use of the secondary dns while the primary is available.
If anyone can shed some light on this, please do.