Core was previously ignoring JSON-encoded dynamic values, but these are
technically supported, so we must either error or accept the value.
Since we already have the decoder for Json state, it's minimal effort to
support this on all plugin methods too.
This change also gives providers an easy way to implement the
UpgradeResourceState method. The obvious implementation of returning the same
JSON-encoded value has tripped up a few providers not using the legacy
SDK already, and we should have at least indicated that the value was
being lost.
A cost estimation error does not actually stop a run, so the run was continuing in the background after the cli exits, causing confusion. This change matches the UI behavior.
For normal provider installation we want to associate each provider with
a selected version number and find a suitable package for that version
that conforms to the official hashes for that release.
Those requirements are very onerous for a provider developer currently
testing a not-yet-released build, though. To allow for that case this new
CLI configuration feature allows overriding specific providers to refer
to give local filesystem directories.
Any provider overridden in this way is not subject to the usual
restrictions about selected versions or checksum conformance, and
activating an override won't cause any changes to the selections recorded
in the lock file because it's intended to be a temporary setting for one
developer only.
This is, in a sense, a spiritual successor of an old capability we had to
override specific plugins in the CLI configuration file. There were
some vestiges of that left in the main package and CLI config package
but nothing has actually been honoring them for several versions now and
so this commit removes them to avoid confusion with the new mechanism.
Core was previously ignoring JSON-encoded dynamic values, but these are
technically supported, so we must either error or accept the value.
Since we already have the decoder for Json state, it's minimal effort to
support this on all plugin methods too.
This change also gives providers an easy way to implement the
UpgradeResourceState method. The obvious implementation of returning the same
JSON-encoded value has tripped up a few providers not using the legacy
SDK already, and we should have at least indicated that the value was
being lost.
If the provisioner configuration includes sensitive values, it's a
reasonable assumption that we should suppress its log output. Obvious
examples where this makes sense include echoing a secret to a file using
local-exec or remote-exec.
This commit adds tests for both logging output from provisioners with
non-sensitive configuration, and suppressing logs for provisioners with
sensitive values in configuration.
Note that we do not suppress logs if connection info contains sensitive
information, as provisioners should not be logging connection
information under any circumstances.
If provisioner configuration or connection info includes sensitive
values, we need to unmark them before calling the provisioner. Failing
to do so causes serialization to error.
Unlike resources, we do not need to capture marked paths here, so we
just discard the marks.
If a NestingList or NestingMap contains a dynamic type, they must be
handled as a cty.Tuple and cty.Object respectively, because the elements
may not have precisely matching types.
In this case, "atomic" means that there will be no situation where the
file contains only part of the newContent data, and therefore other
software monitoring the file for changes (using a mechanism like inotify)
won't encounter a truncated file.
It does _not_ mean that there can't be existing filehandles open against
the old version of the file. On Windows systems the write will fail in
that case, but on Unix systems the write will typically succeed but leave
the existing filehandles still pointing at the old version of the file.
They'll need to reopen the file in order to see the new content.
This originated in the cliconfig code to write out credentials files. The
Windows implementation of this in particular was quite onerous to get
right because it needs a very specific sequence of operations to avoid
running into exclusive file locks, and so by factoring this out with
only cosmetic modification we can avoid repeating all of that engineering
effort for other atomic file writing use-cases.
Terraform v0.10 introduced .terraform/plugins as a cache directory for
automatically-installed plugins, Terraform v0.13 later reorganized the
directory structure inside but retained its purpose as a cache.
The local cache used to also serve as a record of specifically which
packages were selected in a particular working directory, with the intent
that a second run of "terraform init" would always select the same
packages again. That meant that in some sense it behaved a bit like a
local filesystem mirror directory, even though that wasn't its intended
purpose.
Due to some unfortunate miscommunications, somewhere a long the line we
published some documentation that _recommended_ using the cache directory
as if it were a filesystem mirror directory when working with Terraform
Cloud. That was really only working as an accident of implementation
details, and Terraform v0.14 is now going to break that because the source
of record for the currently-selected provider versions is now the
public-facing dependency lock file rather than the contents of an existing
local cache directory on disk.
After some consideration of how to move forward here, this commit
implements a compromise that tries to avoid silently doing anything
surprising while still giving useful guidance to folks who were previously
using the unsupported strategy. Specifically:
- The local cache directory will now be .terraform/providers rather than
.terraform/plugins, because .terraform/plugins is effectively "poisoned"
by the incorrect usage that we can't reliably distinguish from prior
version correct usage.
- The .terraform/plugins directory is now the "legacy cache directory". It
is intentionally _not_ now a filesystem mirror directory, because that
would risk incorrectly interpreting providers automatically installed
by Terraform v0.13 as if they were a local mirror, and thus upgrades
and checksum fetches from the origin registry would be blocked.
- Because of the previous two points, someone who _was_ trying to use the
legacy cache directory as a filesystem mirror would see installation
fail for any providers they manually added to the legacy directory.
To avoid leaving that user stumped as to what went wrong, there's a
heuristic for the case where a non-official provider fails installation
and yet we can see it in the legacy cache directory. If that heuristic
matches then we'll produce a warning message hinting to move the
provider under the terraform.d/plugins directory, which is a _correct_
location for "bundled" provider plugins that belong only to a single
configuration (as opposed to being installed globally on a system).
This does unfortunately mean that anyone who was following the
incorrectly-documented pattern will now encounter an error (and the
aforementioned warning hint) after upgrading to Terraform v0.14. This
seems like the safest compromise because Terraform can't automatically
infer the intent of files it finds in .terraform/plugins in order to
decide automatically how best to handle them.
The internals of the .terraform directory are always considered
implementation detail for a particular Terraform version and so switching
to a new directory for the _actual_ cache directory fits within our usual
set of guarantees, though it's definitely non-ideal in isolation but okay
when taken in the broader context of this problem, where the alternative
would be silent misbehavior when upgrading.
Alisdair McDiarmid