aws-vpc-move-ip: Don't warn for expected scenarios

Make log levels more appropriate to the situation. Before this patch, a
normal start looked like this:

Operation start for aws-vip (ocf:heartbeat:aws-vpc-move-ip) returned: 'ok' (0)
> stderr: Nov 09 02:38:20 INFO: EC2: Moving IP address 10.0.1.65 to this host by adjusting routing table rtb-01b4ea1ae0ec0a4d9
> stderr: Nov 09 02:38:20 INFO: monitor: check routing table (API call) - rtb-01b4ea1ae0ec0a4d9
> stderr: Nov 09 02:38:22 WARNING: IP 10.0.1.65 not assigned to running interface
> stderr: Nov 09 02:38:22 INFO: EC2: Adjusting routing table and locally configuring IP address
> stderr: RTNETLINK answers: Cannot assign requested address
> stderr: Nov 09 02:38:24 WARNING: command failed, rc 2
> stderr: Nov 09 02:38:24 INFO: monitor: check routing table (API call) - rtb-01b4ea1ae0ec0a4d9

Now it looks like this:

Operation start for aws-vip (ocf:heartbeat:aws-vpc-move-ip) returned: 'ok' (0)
> stderr: Nov 09 02:40:43 INFO: EC2: Moving IP address 10.0.1.65 to this host by adjusting routing table rtb-01b4ea1ae0ec0a4d9
> stderr: Nov 09 02:40:43 INFO: monitor: check routing table (API call) - rtb-01b4ea1ae0ec0a4d9
> stderr: Nov 09 02:40:44 INFO: IP 10.0.1.65 not assigned to running interface
> stderr: Nov 09 02:40:44 INFO: EC2: Adjusting routing table and locally configuring IP address
> stderr: Nov 09 02:40:46 INFO: monitor: check routing table (API call) - rtb-01b4ea1ae0ec0a4d9

Under normal circumstances, the call to ec2ip_drop() within
ec2ip_get_and_configure should not be required at all. The drop
function deletes the address before the get_and_configure function
immediately re-adds the address. This call could probably be removed
altogether. Instead, I left the call and silenced its output just in
case of unexpected edge cases.

Resolves: RHBZ#1895811

Signed-off-by: Reid Wahl <nrwahl@protonmail.com>