When using instance templates, if you use one of our image shorthands at
the moment, you'll get a perma-diff. This is because the config gets
resolved to another format before it is set in state, and so we need to
set that other format in state.
Unfortunately, resolving images requires network access, so we have to
do this with CustomizeDiff. CustomizeDiff was having trouble (I think?
More on this below) on setting a field as not ForceNew once the field
was already set, so I moved the decision for whether a field was
ForceNew or not into CustomizeDiff. I also resolved the old and new
images, and if they were the same, cleared the diff.
Unfortunately, you can't actually clear a field on a sub-block right
now. You have to clear top-level fields only. So this will currently
throw an error. I opened hashicorp/terraform#18795 to fix that. Once
that's merged, and we vendor it here, this patch fixes the problem.
If hashicorp/terraform#18795 doesn't get merged, the next best
workaround is to keep track of _all_ the fields under `disk` with a diff
in our CustomizeDiff, check whether they've all changed or not, and if
they've all changed, clear the changes on `disk`, which I _think_ will
resolve the issue. That's just a massive pain, unfortunately.
* fix service account key data source name
* switch id to name
* update docs
* doc format
* fixes for validation and tests
* last fixes for service account key data source
Previously, we were only setting source images for the disks in instance
templates if they weren't set in the config. Instead, let's just always
set them. This fixes our test failures about self_links not matching
names. Also, relative links are safer than just image names, as
Terraform wouldn't notice if another project's image of the same name
got used instead of your project's.
Also fix the setting of tags and tag fingerprints to always set, even if
to the empty value, to fix the tests.
* test empty authorized masters' cidr_blocks
When the `cidr_block` isn't simply blank but contains an empty list as in
```
master_authorized_networks_config {
cidr_blocks = []
}
```
a panic occurs looking something like
```
goroutine 26 [running]:
github.com/terraform-providers/terraform-provider-google/google.expandMasterAuthorizedNetworksConfig(0x15a4f80, 0xc4202586e0, 0x21)
/tmp/GOPATH/src/github.com/terraform-providers/terraform-provider-google/google/resource_container_cluster.go:1355 +0x4f2
github.com/terraform-providers/terraform-provider-google/google.resourceContainerClusterCreate(0xc420146a80, 0x16b1800, 0xc4200b8000, 0x0, 0x0)
/tmp/GOPATH/src/github.com/terraform-providers/terraform-provider-google/google/resource_container_cluster.go:520 +0x2848
github.com/terraform-providers/terraform-provider-google/vendor/github.com/hashicorp/terraform/helper/schema.(*Resource).Apply(0xc420495490, 0xc420341310, 0xc4202582c0, 0x16b1800, 0xc4200b8000, 0x1, 0xc42024eae0, 0xc4201e3650)
/tmp/GOPATH/src/github.com/terraform-providers/terraform-provider-google/vendor/github.com/hashicorp/terraform/helper/schema/resource.go:227 +0x364
github.com/terraform-providers/terraform-provider-google/vendor/github.com/hashicorp/terraform/helper/schema.(*Provider).Apply(0xc4204c6700, 0xc4203412c0, 0xc420341310, 0xc4202582c0, 0x14ee1441a000, 0x0, 0x18)
/tmp/GOPATH/src/github.com/terraform-providers/terraform-provider-google/vendor/github.com/hashicorp/terraform/helper/schema/provider.go:283 +0xa4
github.com/terraform-providers/terraform-provider-google/vendor/github.com/hashicorp/terraform/plugin.(*ResourceProviderServer).Apply(0xc4202d7c40, 0xc42035de80, 0xc42025c160, 0x0, 0x0)
/tmp/GOPATH/src/github.com/terraform-providers/terraform-provider-google/vendor/github.com/hashicorp/terraform/plugin/resource_provider.go:527 +0x57
reflect.Value.call(0xc4203feae0, 0xc42000e038, 0x13, 0x19e88a8, 0x4, 0xc42015ff20, 0x3, 0x3, 0xc420047ee8, 0xc4204c6798, ...)
/usr/local/go/src/reflect/value.go:434 +0x905
reflect.Value.Call(0xc4203feae0, 0xc42000e038, 0x13, 0xc420047f20, 0x3, 0x3, 0xc400000001, 0x0, 0x0)
/usr/local/go/src/reflect/value.go:302 +0xa4
net/rpc.(*service).call(0xc420418600, 0xc42007c140, 0xc42001e798, 0xc4200c4000, 0xc4202d6c40, 0x1557f80, 0xc42035de80, 0x16, 0x1557fc0, 0xc42025c160, ...)
/usr/local/go/src/net/rpc/server.go:381 +0x142
created by net/rpc.(*Server).ServeCodec
/usr/local/go/src/net/rpc/server.go:475 +0x36b
```
which we trigger by altering the first step to contain the HCL notation
for an empty list instead of simply an empty string.
In order to accomplish this, the tests had to be modified to accept an
emptyValue string as well which contains the content of the
`emptyValue` string when the cidrBlocks array is empty. This maintains
the old behavior of the original tests when `emptyValue` is an empty
string, while also facilating differing behavior for the new testcase by
setting `emptyValue` to whichever string we want to test instead. I
don't think this is very clean, but I guess it's pragmatic enough.
I'll hear if this is a thorn in the side to someone 😏.
* avoid panic on cidr_block type assertion
This is basically the fix. Since the value can be nil, we want to ensure
we handle a failure during the assertion since we know that asserting
`nil` conforms to `map[string]interface{}` will cause a run-time panic.
* flatten to config on empty list for cidr_blocks
since an empty list for cidrBlocks constitutes valid input, one should
return a map containing an empty list for the cidr_blocks field instead
of a nil value.
The nil value is only appropriate when the input Config is also nil.
* Update google_container_cluster master_auth username and password without recreation of cluster, using setMasterAuth method
* Add tests to update and disable master_auth password and username values