Unignore empty values in the provider configuration block in 5.0.0
#9014
Conversation
This comment was marked as resolved.
This comment was marked as resolved.
SarahFrench
changed the base branch from
main
to
FEATURE-BRANCH-major-release-5.0.0
|
/gcbrun |
|
Hi there, I'm the Modular magician. I've detected the following information about your changes: Diff reportYour PR generated some diffs in downstreams - here they are. Terraform GA: Diff ( 1 file changed, 1 insertion(+), 2 deletions(-)) |
Tests analyticsTotal tests: Action takenFound 1 affected test(s) by replaying old test recordings. Starting RECORDING based on the most recent commit. Click here to see the affected testsTestAccVertexAIIndexEndpoint_updated |
|
|
SarahFrench
force-pushed
the
unignore-empty-vales-in-5.0.0
branch
from
a74069f to
1ea56a2
Compare
|
Hi there, I'm the Modular magician. I've detected the following information about your changes: Diff reportYour PR generated some diffs in downstreams - here they are. Terraform GA: Diff ( 1 file changed, 1 insertion(+), 2 deletions(-)) |
Tests analyticsTotal tests: Action takenFound 6 affected test(s) by replaying old test recordings. Starting RECORDING based on the most recent commit. Click here to see the affected testsTestAccAlloydbInstance_createInstanceWithNetworkConfigAndAllocatedIPRange|TestAccBigQueryDataTable_bigtable|TestAccBigtableAppProfile_bigtableAppProfileAnyclusterExample|TestAccDataprocJobIamPolicy|TestAccSpannerDatabaseIamPolicy|TestAccVertexAIIndexEndpoint_updated |
|
Rerun these tests in REPLAYING mode to catch issues
|
SarahFrench
changed the title
provider configuration block in 5.0.0
|
Hi there, I'm the Modular magician. I've detected the following information about your changes: Diff reportYour PR generated some diffs in downstreams - here they are. Terraform GA: Diff ( 4 files changed, 58 insertions(+), 138 deletions(-)) |
|
Looks like there's a unit test failure! Also, is it surprising that we generated a TC config in TPG but not TPGB? |
Ah, thanks for pointing it out!
Yeah, very surprising 🤔 It's a sign that the packages.kt file isn't being generated properly when the downstream is made- I'll look into it Edit: I can't see any evidence of that generated/packages.kt file being changed unexpectedly in the TPG/TPGB repos, so I think think there's a real problem. That file is generated by iterating through package folders, so it could be that there's a bit of a race condition? Edit 2: this might be something that's specific to contributing to the 5.0.0 release - I see the same here #8587 (comment) |
Tests analyticsTotal tests: Action takenFound 3 affected test(s) by replaying old test recordings. Starting RECORDING based on the most recent commit. Click here to see the affected testsTestAccAlloydbInstance_createInstanceWithNetworkConfigAndAllocatedIPRange|TestAccSpannerDatabaseIamPolicy|TestAccVertexAIIndexEndpoint_updated |
|
Rerun these tests in REPLAYING mode to catch issues
|
…nguish between `""` in config vs zero values
|
Hi there, I'm the Modular magician. I've detected the following information about your changes: Diff reportYour PR generated some diffs in downstreams - here they are. Terraform GA: Diff ( 6 files changed, 60 insertions(+), 140 deletions(-)) |
SarahFrench
changed the title
provider configuration block in 5.0.0provider configuration block in 5.0.0
Tests analyticsTotal tests: Action takenFound 4 affected test(s) by replaying old test recordings. Starting RECORDING based on the most recent commit. Click here to see the affected testsTestAccAlloydbInstance_createInstanceWithNetworkConfigAndAllocatedIPRange|TestAccBigtableAppProfile_bigtableAppProfileSingleclusterExample|TestAccBigtableAppProfile_bigtableAppProfileMulticlusterExample|TestAccVertexAIIndexEndpoint_updated |
|
Rerun these tests in REPLAYING mode to catch issues
|
|
Ack, I forgot the guide - I'll make the PR for that now |
First PR to address hashicorp/terraform-provider-google#14447
This PR stops the provider ignoring empty strings in provider blocks; if the value is in a user's config then it should be used by the code or removed by the user.
After muxing the provider we accidentally changed its behaviour to stop ignoring empty strings. That behaviour is what we'd expect the correct behaviour to be, so we are reverting to the old behaviour in a 4.x.x release (see #8798) and then intentionally making the switch to the new behaviour as part of 5.0.0
ALSO I've decided against making similar changes in #9015 because it would result in breaking changes for users when any datasources/resources are migrated to the plugin framework.
To do:
Release Note Template for Downstream PRs (will be copied)