bad3a876ca
provider/cloudstack: add support for multiple NICs with port forwards and set network_domain for networks ( #10638 )
...
* Add support for multiple NICs with port forwards
* Fix issue #9801
2016-12-12 10:06:42 +01:00
15cd542392
Remove a duplicate `connect_timeout` from a rebase+stash.
2016-12-12 00:53:06 -08:00
5016a56fd4
provider/statuscake: fix StatusCake tests ( #10660 )
...
* Update vendored statuscake SDK
* Set all attributes when upserting statuscake tests
2016-12-11 19:41:49 +00:00
42edd22566
helper/variables: address go vet
2016-12-10 20:32:50 -05:00
b346ba32d1
terraform: dependent provider resources are destroyed first in modules
...
This extends the prior commit to also verify (and fix) that resources of
dependent providers are destroyed first even when they're within
modules.
2016-12-10 20:22:12 -05:00
14d079f914
terraform: destroy resources in dependent providers first
...
Fixes #4645
This is something that never worked (even in legacy graphs), but as we
push forward towards encouraging multi-provider usage especially with
things like the Vault data source, I want to make sure we have this
right for 0.8.
When you have a config like this:
```
resource "foo_type" "name" {}
provider "bar" { attr = "${foo_type.name.value}" }
resource "bar_type" "name" {}
```
Then the destruction ordering MUST be:
1. `bar_type`
2. `foo_type`
Since configuring the client for `bar_type` requires accessing data from
`foo_type`. Prior to this PR, these two would be done in parallel. This
properly pushes forward the dependency.
There are more cases I want to test but this is a basic case that is
fixed.
2016-12-10 20:11:24 -05:00
3ba9720b3e
config: validate invalid variable keys
...
Fixes #9416
A simple change to verify that only valid keys for `variable` blocks are
used.
2016-12-10 19:27:01 -05:00
cabcc4b5b8
terraform: allow indexing into a computed list for multi-count resources
...
Fixes #8695
When a list count was computed in a multi-resource access
(foo.bar.*.list), we were returning the value as empty string. I don't
actually know the histocal reasoning for this but this can't be correct:
we must return unknown.
When changing this to unknown, the new tests passed and none of the old
tests failed. This leads me further to believe that the return empty
string is probably a holdover from long ago to just avoid crashes or
UUIDs in the plan output and not actually the correct behavior.
2016-12-10 19:17:29 -05:00
5d684b399c
config: error when loading multiple lifecycle blocks
...
Fixes #8776
This introduces an error when multiple `lifecycle` blocks exist on a
resource in the configuration.
2016-12-10 18:52:13 -05:00
c02b2471d6
Fixed typoes in Consul key_prefix docs ( #10655 )
2016-12-10 21:45:47 +00:00
da418b2582
website: note about map/var merging
2016-12-10 15:05:58 -05:00
83e72b0361
helper/variables: ParseInput for consistent parsing
2016-12-10 14:44:17 -05:00
77efacf30e
command: use helper/variables for flags and parsing
2016-12-10 14:30:40 -05:00
39e3d8ea9b
helper/variables: helpers for working with vars
2016-12-10 14:30:36 -05:00
e6b1cda7db
documentation/aws: Fix example syntax ( #10651 )
2016-12-10 09:36:06 +00:00
e2448473cb
Clean up the DatabaseCreate call.
2016-12-10 12:32:21 +11:00
db5d7b0438
Style nit
2016-12-10 12:32:21 +11:00
bfc2a2d42f
Commit WIP re: updated postgresql_role provider.
...
*Read() and *Update() still need to be updated.
2016-12-10 12:32:20 +11:00
2e529146a5
Remove unused code.
2016-12-10 12:32:20 +11:00
6b540ecb55
Don't use d.GetOk() when the zero value is a required attribute.
...
Add "pathological" test.
2016-12-10 12:32:20 +11:00
37fdc958b3
Always remove the IS_TEMPLATE attribute before dropping a database.
2016-12-10 12:32:20 +11:00
e9b2b38288
Remove SetId() call from *Read(), this isn't required for import to work.
2016-12-10 12:32:20 +11:00
f253fc9eff
Commit miss earlier, fix syntax from unstaged commit.
2016-12-10 12:32:20 +11:00
5280c37bea
`postgresql_database` resource provider should now be feature complete.
...
* Add support to import databases. See docs.
* Add support for renaming databases
* Add support for all known PostgreSQL database attributes, including:
* "allow_connections"
* "lc_ctype"
* "lc_collate"
* "connection_limit"
* "encoding"
* "is_template"
* "owner"
* "tablespace_name"
* "template"
2016-12-10 12:32:20 +11:00
59f4ad6fd1
Alpha sort the resources in the PostgreSQL map.
2016-12-10 12:32:19 +11:00
300ef2bc97
Add `connect_timeout` support to the PostgreSQL provider.
2016-12-10 12:32:19 +11:00
3750bf7af2
Depreciate the PostgreSQL provider's `ssl_mode` option in favor of `sslmode`.
...
Both libpq(3) and github.com/lib/pq both use `sslmode`. Prefer this vs
the non-standard `ssl_mode`. `ssl_mode` is supported for compatibility
but should be removed in the future.
Changelog: yes
2016-12-10 12:32:19 +11:00
a200899d93
Allow the PostgreSQL provider to talk to a non-default database.
2016-12-10 12:32:19 +11:00
602a908d7b
Add a `fallback_application_name` to the PostgreSQL DSN.
2016-12-10 12:32:19 +11:00
44d907a3de
Revert "Prefer PGHOSTADDR over PGHOST, if set." lib/pq doesn't support PGHOSTADDR.
...
d8eeeb8bae/conn.go (L1851)
2016-12-10 12:32:19 +11:00
55061d190a
Prefer PGHOSTADDR over PGHOST, if set.
2016-12-10 12:32:19 +11:00
3779dfffa9
Use a string instead of the `%t` modifier for printing a bool in SQL
2016-12-10 12:32:19 +11:00
5b66bf0745
ForceNew is required when changing the locale, ctype, and encoding.
...
This will cause someone some grief. TODO: Figure out how to prevent
someone from blowing off their foot if they twiddle this after the
fact.
2016-12-10 12:32:18 +11:00
547dcf27e1
Decorate the computed attribute where appropriate
2016-12-10 12:32:18 +11:00
242405bdf1
Factor out the validate function for connection limits
2016-12-10 12:32:18 +11:00
02dea2edd9
Add missing descriptions to owner and name
2016-12-10 12:32:18 +11:00
deb91f61ce
Update the descriptions to mostly match the official PostgreSQL docs.
2016-12-10 12:32:18 +11:00
68cadd3c2a
Update the FIXME/TODO style to match Google style guides
2016-12-10 12:32:18 +11:00
f3add9e7ef
Flesh out the CREATE DATABASE for PostgreSQL.
2016-12-10 12:32:18 +11:00
2aee081e4b
Eh, specify default values lib/pq zero initializes itself
2016-12-10 12:32:17 +11:00
7e5ccc089b
Spell `ssl_mode` like `sslmode`
2016-12-10 12:32:17 +11:00
655617ed6a
Remove the PGPASSWORD requirement for tests.
...
If they're required and the value is missing, the test will fail. There's
no need to enforce that in the test itself.
2016-12-10 12:32:17 +11:00
ec130d538c
Add comments requesting that the testing-specific environment variables
...
can be removed in Terraform 0.8.0
2016-12-10 12:32:17 +11:00
b68ef2c40b
Fall through to using the defaults from github.com/lib/pq
2016-12-10 12:32:17 +11:00
f31ebff10e
Change the PostgreSQL PGSSLMODE option to sslmode to match PostgreSQL idioms.
...
Also don't specify the default and rely on github.com/lib/pq (which uses "required"
and is different than what libpq(3) uses, which is "preferred" and unsupported by
github.com/lib/pq).
2016-12-10 12:32:17 +11:00
df7a15fff9
Complete the docs for Consul Prepared Query.
...
While here add an additional example of a standard prepared query in
addition to the prepared query template that was already present.
2016-12-10 12:28:06 +11:00
808f09f01f
terraform: user friendly error when using old map overrides
...
Related to #8036
We have had this behavior for a _long_ time now (since 0.7.0) but it
seems people are still periodically getting bit by it. This adds an
explicit error message that explains that this kind of override isn't
allowed anymore.
2016-12-09 15:58:24 -05:00
1f5b940f3f
Merge pull request #10175 from hashicorp/b-aws-elasticache-updates
...
[WIP] provider/aws: Update Elasticache docs for corrections, clarity
2016-12-09 14:56:26 -06:00
626ad57546
Merge pull request #10643 from hashicorp/jbardin/GH-10640
...
Disaply interpolation strings for computed fields
2016-12-09 15:31:42 -05:00
4a38f5f534
Merge pull request #10642 from hashicorp/b-num-id
...
config: allow names to start with numbers
2016-12-09 14:42:21 -05:00
Sander van Harmelen