I have set up the nameserver for my VPS service using bind9. My server has Ubuntu 10.04LTS setup.
The problem that I am facing is that, for some people, they cannot connect to my server at all. I think it has to do something with DNS settings.
To test that problem, I did the following test. Lets say my server as server A (nameserver).
From random server B, if I dig to one of domain hosted on my server A it always finds returns the 'NoError' result. (which means successful)
However, for random server C, if I dig to one of domain hosted on my server, about half of the time, it returns SERVFAIL result and the other half 'NoError'. This is really really weird, because how can dig fail while it was successful like few seconds ago?
One of domains using my nameserver (bind) is like below.
$TTL 604800
@ IN SOA ns.xxxx.net. admin.xxxx.net. (
20110120 ; Serial
604800 ; Refresh
86400 ; Retry
2419200 ; Expire
604800 ) ; Negative Cache TTL
;
@ IN NS ns.xxx.net.
@ IN A xxx.xxx.xxx.78
ns IN A xxx.xxx.xxx.78
www IN A xxx.xxx.xxx.78
Is there anything exceptional that may cause problem? Some people say TTL 604800 is just too long.. is it?