CNAME resolution of AWS Aurora database cluster address

### `consul version` for both Client and Server
Client: `v0.8.2` and `v0.8.3`
Server: `v0.8.2` and `v0.8.3`

### Operating system and Environment details
CentOS7 x86_64

### Description of the Issue (and unexpected/desired result)

Originally posted in the [mailing list](https://groups.google.com/d/msg/consul-tool/x-eceQxHP-U/REiH8IbOBAAJ).

The Aurora database service provides an address for each specific database instance and a _cluster address_ (a CNAME pointing to the current database leader/writer).

Resolution of this cluster address through Consul stops with a CNAME instead of an A record.

This is on an otherwise working Consul cluster using dnsmasq and the AWS VPC DNS recursor.

### Reproduction steps

1. Launch a VPC in AWS
2. Launch an [Aurora RDS cluster](https://aws.amazon.com/rds/aurora/)
3. Launch an EC instance and configure consul and dnsmasq
4. Register 2 services with Consul catalog:

A "db" service in Consul configured with the **instance** address/endpoint:
```
; <<>> DiG 9.8.2rc1-RedHat-9.8.2-0.37.rc1.43.amzn1 <<>> db.service.consul
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 487
;; flags: qr aa rd ra; QUERY: 1, ANSWER: 3, AUTHORITY: 0, ADDITIONAL: 0

;; QUESTION SECTION:
;db.service.consul. IN A

;; ANSWER SECTION:
db.service.consul. 5 IN CNAME db.XXXXXXXXXXXX.us-west-2.rds.amazonaws.com.
db.XXXXXXXXXXXX.us-west-2.rds.amazonaws.com. 5 IN CNAME ec2-XXX-XXX-XXX-XXX.us-west-2.compute.amazonaws.com.
ec2-XXX-XXX-XXX-XXX.us-west-2.compute.amazonaws.com. 20 IN A 10.XXX.XXX.XXX
```

A "db" service in Consul configured with the **cluster** address/endpoint:
```
; <<>> DiG 9.8.2rc1-RedHat-9.8.2-0.37.rc1.43.amzn1 <<>> db.service.consul
;; global options: +cmd
;; Got answer:
;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 48735
;; flags: qr aa rd ra; QUERY: 1, ANSWER: 3, AUTHORITY: 0, ADDITIONAL: 0

;; QUESTION SECTION:
;db.service.consul. IN A

;; ANSWER SECTION:
db.service.consul. 5 IN CNAME db-cluster.cluster-XXXXXXXXXXXX.us-west-2.rds.amazonaws.com.
db-cluster.cluster-XXXXXXXXXXXX.us-west-2.rds.amazonaws.com. 5 IN CNAME db.XXXXXXXXXXXX.us-west-2.rds.amazonaws.com.
db.XXXXXXXXXXXX.us-west-2.rds.amazonaws.com. 5 IN CNAME ec2-XXX-XXX-XXX-XXX.us-west-2.compute.amazonaws.com.
```

### Expected result:
Clients can connect via the consul dns using either configuration.

### Actual result:
Clients can connect via the consul dns using the first configuration.

Clients cannot connect via the consul dns using this second configuration due to the missing A record.

### Workaround:

Using the first configuration works for now, but misses out on the benefit of automatic failover of the leader/writer instance.

The Aurora database service also offers a cluster "reader" address/endpoint that provides a CNAME over a pool of read-only instances. While I have not tried it, I suspect it will have DNS resolution issues as the writer address.