kubernetes: Ignore internal K8S annotations in config_map + use PATCH

```
% make testacc TEST=./builtin/providers/vsphere TESTARGS='-run=TestAccVSphereVirtualDisk_'
==> Checking that code complies with gofmt requirements...
go generate $(go list ./... | grep -v /terraform/vendor/)
2017/03/27 18:08:08 Generated command/internal_plugin_list.go
TF_ACC=1 go test ./builtin/providers/vsphere -v -run=TestAccVSphereVirtualDisk_ -timeout 120m
=== RUN   TestAccVSphereVirtualDisk_basic
--- PASS: TestAccVSphereVirtualDisk_basic (8.57s)
PASS
ok  	github.com/hashicorp/terraform/builtin/providers/vsphere	8.591s
```

Merge branch 'vsphere-disk-capacity' of https://github.com/peay/terraform into peay-vsphere-disk-capacity

In function `resourceAwsSesReceiptRuleRead` the position of the receipt
rules in the rule set was taken directly from the index of the rule's
position in the slice returned by the AWS API call. As the slice is
zero-based and the ruleset is one-based, this results in an incorrect
representation.

This manifests as `aws_ses_receipt_rule` resources always showing a
diff during plan or apply.

docs: Update getting started docs to use a valid AMI in us-east-1, vpc

Allow to resize a droplet permanently (i.e. apply disk resize)
if previously it was resized temporarily (CPU and RAM only).

Fixes: #13007

```
$ make testacc TEST=./builtin/providers/digitalocean TESTARGS='-run=TestAccDigitalOceanDroplet_ResizeOnlyDisk'
==> Checking that code complies with gofmt requirements...
go generate $(go list ./... | grep -v /terraform/vendor/)
2017/03/25 03:54:23 Generated command/internal_plugin_list.go
TF_ACC=1 go test ./builtin/providers/digitalocean -v -run=TestAccDigitalOceanDroplet_ResizeOnlyDisk -timeout 120m
=== RUN   TestAccDigitalOceanDroplet_ResizeOnlyDisk
--- PASS: TestAccDigitalOceanDroplet_ResizeOnlyDisk (198.62s)
PASS
ok  	github.com/hashicorp/terraform/builtin/providers/digitalocean	198.638s
```

according to the official scaleway support, requests within the same session can
not be parallelized.

While I do not know for sure that this is a write-only limitation, I've
implemented it as a write-only limitation for now.

Previously requests like this would produce a 500 internal server error:

```
resource "scaleway_ip" "test_ip" {
  count = 2
}
```

now this limitation should be lifted, for all scaleway resources

* vendor: adding golang-petname for random_pet resource

* provider/random: random_pet resource

This commit accounts for a 409 response when a LBaaS v2 pool is
being created. Rather than error out, this should be considered a
pending state.

The OS_DEBUG feature has worked out great, but frequent logging of
the service catalog during client initialization can make logging
very chatty. This commit omits the service catalog in the logs.

This commit deprecates the floating_ip attributes from the
openstack_compute_instance_v2 resource. It is recommended to use
either the openstack_compute_floatingip_associate resource or
configure an openstack_networking_port_v2 resource with a floating
IP.

This commit deprecates the volume attribute in the
openstack_compute_instance_v2 resource. It's recommended to either
use the block_device attribute or the openstack_compute_volume_attach_v2
resource.

provider/openstack: Fixing typo in secgroup rule timeout test

* Add the `consul` check type to the `circonus_check` resource.

* Fix a tab-complete fail.

`Parse` != `Path`, but lexically close.

* Dept of 2nd thoughts: `s/service_name/service/g`

Supersedes #12932 , populating the entire guide with the correct AMI(s)

* fix bool vs Compatobool value used in flattendBackends

* remove log line

* update test case