The provisionerFail_createBeforeDestroy test was verifying the incorrect
output. The create_before_destroy instance in the state has an ID of
"bar" with require_new="abc", and a new instance would get an ID of
"foo" with require_new="xyz". The existing test was expecting the
following state:
aws_instance.bar: (1 deposed)
ID = bar
provider = provider.aws
require_new = abc
Deposed ID 1 = foo (tainted)
Which showed "bar" still the primary instance in the state, with the new
instance "foo" as being the deposed instance, though properly tainted.
The new output is:
aws_instance.bar: (tainted) (1 deposed)
ID = foo
provider = provider.aws
require_new = xyz
type = aws_instance
Deposed ID 1 = bar
Showing the new "foo instance as being the primary instance in the
state, with "bar" as the deposed instance.
Use the new ExitStatus method, and also check the cmd.Err() method for
errors.
Remove leaks from the output goroutines in both provisioners by
deferring their cleanup, and returning early on all error conditions.
This was updated to see if we can get at any error status from the
remote command and transport, which still is not available, but kept the
latest version since it fixes a couple race conditions.
The timeout for the remote command was taken from the wrong config
field, and the connection timeout was being used which is 5 min. Any
remote command taking more than 5 min would be terminated by
disconnecting the communicator. Remove the timeout from the context, and
rely on the global timeout provided by terraform.
There was no way to get the error from the communicator previously, so
the broken connection was silently ignored and the provisioner returned
successfully. Now we can use the new cmd.Err() method to retrieve any
errors encountered during execution.
The remote.Cmd struct could not convey any transport related error to
the caller, meaning that interrupted commands would show that they
succeeded.
Change Cmd.SetExited to accept an exit status, as well as an error to
store for the caller. Make the status and error fields internal,
require serialized access through the getter methods.
Users of remote.Cmd should not check both Cmd.Err() and Cmd.ExitStatus()
until after Wait returns.
Require communicators to call Cmd.Init before executing the command.
This will indicate incorrect usage of the remote.Cmd by causing a panic
in SetExitStatus.
Fixes: #17314
We now deal correctly with the creation of the state file - we were
not dealing well with a ResourceNotFound error
Now that this has been changed around, we try and create the statefile
and if there is an error, we look for an existing statefile - previously
this was not the order of operations
If a count field references another count field which is interpolated
but is attached to a resource already in the state, the result of that
first interpolation will be lost when a plan is serialized. This is
because the result of the first interpolation is stored directly in the
module config, in an unexported config field.
This is not a general fix for the above situation, which would require
refactoring how counts are handles throughout the config. Ignoring the
error works, because in most cases the count will be properly
handled during the resource's interpolation.
An interpolated count value that is determined during plan, is lost
during plan serialization, causing apply to fail when the interpolation
string can't be evaluated.
Previously we just ported over the simple "string", "list", and "map" type
hint keywords from the old loader, which exist primarily as hints to the
CLI for whether to treat -var=... arguments and environment variables as
literal strings or as HCL expressions.
However, we've been requested before to allow more specific constraints
here because it's generally better UX for a type error to be detected
within an expression in a calling "module" block rather than at some point
deep inside a third-party module.
To allow for more specific constraints, here we use the type constraint
expression syntax defined as an extension within HCL, which uses the
variable and function call syntaxes to represent types rather than values,
like this:
- string
- number
- bool
- list(string)
- list(any)
- list(map(string))
- object({id=string,name=string})
In native HCL syntax this looks like:
variable "foo" {
type = map(string)
}
In JSON, this looks like:
{
"variable": {
"foo": {
"type": "map(string)"
}
}
}
The selection of literal processing or HCL parsing of CLI-set values is
now explicit in the model and separate from the type, though it's still
derived from the type constraint and thus not directly controllable in
configuration.
Since this syntax is more complex than the keywords that replaced it, for
now the simpler keywords are still supported and "list" and "map" are
interpreted as list(any) and map(any) respectively, mimicking how they
were interpreted by Terraform 0.11 and earlier. For the time being our
documentation should continue to recommend these shorthand versions until
we gain more experience with the more-specific type constraints; most
users should just make use of the additional primitive type constraints
this enables: bool and number.
As a result of these more-complete type constraints, we can now type-check
the default value at config load time, which has the nice side-effect of
allowing us to produce a tailored error message if an override file
produces an invalid situation; previously the result was rather confusing
because the error message referred to the original definition of the
variable and not the overridden parts.
Although we do still consider these deprecated for 0.12, we'll defer
actually generating warnings for them until a later minor release so that
module authors can retain their quoted identifiers for a period after 0.12
release for backward-compatibility with Terraform 0.11.
The error-handling behavior of the HCL parser was improved, which causes
the number of diagnostics and the diagnostics messages to be different
in cases where a block-like introduction is given but without any
following body.
This is largely minor bugfixes for issues found since we last updated the
vendoring. There are some new features here that Terraform is not yet
using and thus present little risk.
In particular this includes the HCL-JSON spec change where arrays can now
be used at any level in a block label structure, to allow for preserving
the relative order of blocks.
While not normally possible, manual manipulation of the state and config
can cause us to end up with a nil config in
evalTreeManagedResourceNoState.
Regardless of how it got here, we can't ever assume the Config field is
not nil, and EvalInterpolate happily accepts a nil RawConfig
James Bardin