Improve diagnostic logging for DNS queries

When using Consul as a DNS server the logs don't provide a lot of information about the status of the DNS queries (even when the log mode has been set to debug or trace). The only information offered is the following:

```
2015/06/16 16:58:20 [DEBUG] dns: cname recurse RTT for mycname.example.com. (1.0065ms)
2015/06/16 16:58:20 [DEBUG] dns: request for {mytag.myservice.service.consul. 1 1} (2.0081ms)
```

For diagnostic purposes like trying to find out if a DNS query has succeeded or not it would be useful to have slighty more information, e.g.:
- Has the CNAME recurse actually worked or did it not return a valid response
- Where did the original request come from
- etc.

No source files, tests, or entry points are named. Start by locating the DNS request and CNAME recursion logging paths and compare their debug/trace output with the examples. Done means the diagnostics report whether CNAME recursion succeeded and where the original request came from.

Written by the indexing model from the issue text.