![]() For `terraform destroy`, we currently build up the same graph we do for `plan` and `apply` and we do a walk with a special Diff that says "destroy everything". We have fought the interpolation subsystem time and again through this code path. Beginning in #2775 we gained a new feature to selectively prune out problematic graph nodes. The past chain of destroy fixes I have been involved with (#6557, #6599, #6753) have attempted to massage the "noop" definitions to properly handle the edge cases reported. "Variable is depended on by provider config" is another edge case we add here and try to fix. This dive only makes me more convinced that the whole `terraform destroy` code path needs to be reworked. For now, I went with a "surgical strike" approach to the problem expressed in #7047. I found a couple of issues with the existing Noop and DestroyEdgeInclude logic, especially with regards to flattening, but I'm explicitly ignoring these for now so we can get this particular bug fixed ahead of the 0.7 release. My hope is that we can circle around with a fully specced initiative to refactor `terraform destroy`'s graph to be more state-derived than config-derived. Until then, this fixes #7407 |
||
---|---|---|
.github | ||
builtin | ||
command | ||
communicator | ||
config | ||
contrib | ||
dag | ||
deps | ||
digraph | ||
dot | ||
examples | ||
flatmap | ||
helper | ||
plugin | ||
scripts | ||
state | ||
terraform | ||
test-fixtures | ||
vendor | ||
website | ||
.gitignore | ||
.travis.yml | ||
BUILDING.md | ||
CHANGELOG.md | ||
LICENSE | ||
Makefile | ||
README.md | ||
Vagrantfile | ||
checkpoint.go | ||
commands.go | ||
config.go | ||
config_test.go | ||
config_unix.go | ||
config_windows.go | ||
help.go | ||
main.go | ||
panic.go | ||
version.go |
README.md
Terraform
- Website: http://www.terraform.io
- IRC:
#terraform-tool
on Freenode - Mailing list: Google Groups
Terraform is a tool for building, changing, and versioning infrastructure safely and efficiently. Terraform can manage existing and popular service providers as well as custom in-house solutions.
The key features of Terraform are:
-
Infrastructure as Code: Infrastructure is described using a high-level configuration syntax. This allows a blueprint of your datacenter to be versioned and treated as you would any other code. Additionally, infrastructure can be shared and re-used.
-
Execution Plans: Terraform has a "planning" step where it generates an execution plan. The execution plan shows what Terraform will do when you call apply. This lets you avoid any surprises when Terraform manipulates infrastructure.
-
Resource Graph: Terraform builds a graph of all your resources, and parallelizes the creation and modification of any non-dependent resources. Because of this, Terraform builds infrastructure as efficiently as possible, and operators get insight into dependencies in their infrastructure.
-
Change Automation: Complex changesets can be applied to your infrastructure with minimal human interaction. With the previously mentioned execution plan and resource graph, you know exactly what Terraform will change and in what order, avoiding many possible human errors.
For more information, see the introduction section of the Terraform website.
Getting Started & Documentation
All documentation is available on the Terraform website.
Developing Terraform
If you wish to work on Terraform itself or any of its built-in providers, you'll first need Go installed on your machine (version 1.6+ is required). Alternatively, you can use the Vagrantfile in the root of this repo to stand up a virtual machine with the appropriate dev tooling already set up for you.
For local dev first make sure Go is properly installed, including setting up a GOPATH. You will also need to add $GOPATH/bin
to your $PATH
.
Next, using Git, clone this repository into $GOPATH/src/github.com/hashicorp/terraform
. All the necessary dependencies are either vendored or automatically installed, so you just need to type make
. This will compile the code and then run the tests. If this exits with exit status 0, then everything is working!
$ make
To compile a development version of Terraform and the built-in plugins, run make dev
. This will build everything using gox and put Terraform binaries in the bin
and $GOPATH/bin
folders:
$ make dev
...
$ bin/terraform
...
If you're developing a specific package, you can run tests for just that package by specifying the TEST
variable. For example below, onlyterraform
package tests will be run.
$ make test TEST=./terraform
...
If you're working on a specific provider and only wish to rebuild that provider, you can use the plugin-dev
target. For example, to build only the Azure provider:
$ make plugin-dev PLUGIN=provider-azure
If you're working on the core of Terraform, and only wish to rebuild that without rebuilding providers, you can use the core-dev
target. It is important to note that some types of changes may require both core and providers to be rebuilt - for example work on the RPC interface. To build just the core of Terraform:
$ make core-dev
Dependencies
Terraform stores its dependencies under vendor/
, which Go 1.6+ will automatically recognize and load. We use govendor
to manage the vendored dependencies.
If you're developing Terraform, there are a few tasks you might need to perform.
Adding a dependency
If you're adding a dependency, you'll need to vendor it in the same Pull Request as the code that depends on it. You should do this in a separate commit from your code, as makes PR review easier and Git history simpler to read in the future.
To add a dependency:
Assuming your work is on a branch called my-feature-branch
, the steps look like this:
- Add the new package to your GOPATH:
go get github.com/hashicorp/my-project
- Add the new package to your vendor/ directory:
govendor add github.com/hashicorp/my-project/package
- Review the changes in git and commit them.
Updating a dependency
To update a dependency:
- Fetch the dependency:
govendor fetch github.com/hashicorp/my-project
- Review the changes in git and commit them.
Acceptance Tests
Terraform has a comprehensive acceptance test suite covering the built-in providers. Our Contributing Guide includes details about how and when to write and run acceptance tests in order to help contributions get accepted quickly.
Cross Compilation and Building for Distribution
If you wish to cross-compile Terraform for another architecture, you can set the XC_OS
and XC_ARCH
environment variables to values representing the target operating system and architecture before calling make
. The output is placed in the pkg
subdirectory tree both expanded in a directory representing the OS/architecture combination and as a ZIP archive.
For example, to compile 64-bit Linux binaries on Mac OS X Linux, you can run:
$ XC_OS=linux XC_ARCH=amd64 make bin
...
$ file pkg/linux_amd64/terraform
terraform: ELF 64-bit LSB executable, x86-64, version 1 (SYSV), statically linked, not stripped
XC_OS
and XC_ARCH
can be space separated lists representing different combinations of operating system and architecture. For example, to compile for both Linux and Mac OS X, targeting both 32- and 64-bit architectures, you can run:
$ XC_OS="linux darwin" XC_ARCH="386 amd64" make bin
...
$ tree ./pkg/ -P "terraform|*.zip"
./pkg/
├── darwin_386
│ └── terraform
├── darwin_386.zip
├── darwin_amd64
│ └── terraform
├── darwin_amd64.zip
├── linux_386
│ └── terraform
├── linux_386.zip
├── linux_amd64
│ └── terraform
└── linux_amd64.zip
4 directories, 8 files
Note: Cross-compilation uses gox, which requires toolchains to be built with versions of Go prior to 1.5. In order to successfully cross-compile with older versions of Go, you will need to run gox -build-toolchain
before running the commands detailed above.