Hello forum.


One of you applications (running on linux) failed due to a DNS server not being available (it was apparently getting restarted). The linux box is set up with two DNS servers in /etc/resolv.conf:
It was the first server which was down for a short while. The second server was, according to logs on the DNS server itself, up while the first one was down.

So my question is this: Why did the application fail even though the second DNS server was up? My understanding of /etc/resolv.conf is that the first server is queried, and after 5 second of non-responsiveness the second server is queried. If this servers doesn't respond, the whole thing goes back to start for 2 more rounds. These values are, if I remember correctly, the default values.

So what may be the reason why the application didn't query that second server? Any ideas anyone?


- kenneho

Perhaps the time-out of the application was shorter then the 5 seconds?

Well, that's a posibility. I'm not sure if the application uses internal linux tools to perform the lookup, or if it uses it's own tools. In case of the formed, I'd suspect that if "host" or "dig" returns an answer everything is good. In case of the latter, it depends on the application specific timeout.

I'll try to find out which of the alternatives the application is using, and post back here.