Closed agility-adamp closed 2 months ago
I was able to get the Connector to successfully deploy by fixing a few config items I mistakenly added to the Connector. These are the fields I removed from the Connector resource in Terraform, which I think were the reason Terraform borked while it was trying to deploy:
So it seems that instead of reporting the errors I'd made, the FiveTran Terraform Provider errored out with the errors I pasted above.
On a side note, I was able to fix the issue about not being able to sync from ALL_TIME; I had to add the timeframe_months=ALL_TIME
setting to the config
block, on the Connector.
There still might be a problem and I'm a bit unsure if it's related but after I fixed these issues, I still get a Connector in the FiveTran UI that appears to need me to manually "Edit the Connection" and then click the "Save and Test" button.
This is what the revised connector config looks like:
resource "fivetran_connector" "ttd" {
service = "the_trade_desk"
group_id = fivetran_group.ingest_group.id
run_setup_tests = "true"
trust_certificates = "true"
trust_fingerprints = "true"
destination_schema {
name = "the_trade_desk_${var.env}"
}
config {
login = data.google_secret_manager_secret_version.fivetran-login-sbx.secret_data
password = data.google_secret_manager_secret_version.fivetran-password-sbx.secret_data
timeframe_months = "ALL_TIME"
sync_mode = "AllAccounts"
}
depends_on = [
fivetran_destination.ingest_dataset
]
}
There still might be a problem and I'm a bit unsure if it's related but after I fixed these issues, I still get a Connector in the FiveTran UI that appears to need me to manually "Edit the Connection" and then click the "Save and Test" button.
Hi @agility-adamp! Thank you for the detailed feedback.
Could you please share your group_id
value so we could track logs for tests to investigate why connection wasn't fully setup after terraform apply.
As for "extra" fields in connector config
block - it's a known issue. We are thinking about additional pre-validation step to prevent users from defining not-expected values in config
block.
Changing the tag to documentation as if we had a documentation with example config for each connector - it would be easier to configure. We could provide additional layer of validation inside provider to prevent such situations also.
Sounds great and thanks for the feedback. Sorry about the extra non-bug ticket and email noise; I actually thought I was looking at a bug when I filed this because it took me a while to figure out that I read the docs wrong and threw in a bunch of extra fields. Thanks again!
@beevital I believe this is my group_id -> infiltration_bald
Describe the bug I am unable to successfully add a TheTradeDesk Connection. Running
terraform apply ....
throws a Terraform error and the error message let's me know this is a problem with the FiveTran Terraform provider.If I am able to get the Connection to deploy, it appears that only some of the configuration I sent in has worked because I've configured the resources to be
paused=false
, which should mean the resource is not paused. It's always paused when I am successful in getting the resource to half deploy (I've never had a fully successful deploy of the FiveTran Connection). The Destination and Connection "connection test"s report a successful test (when I manually trigger the test), so the credentials and IAM role permissions seem to be set up correctly.I've tried changing most of the configuration items in the Connection and the Destination, to no avail.
To Reproduce Properly configure and create a TheTradeDesk Connection using Terraform.I can't find a way to make it work successfully, so I think all you have to do to reproduce this is to attempt to create a FiveTran TheTradeDesk Connection, using Terraform.
This is my config for the Connection and its Destination and Destination Group...
Expected behavior I expected the Connection to be successfully created, unpaused and running its initial synch or at least to have begun it's synch schedule in an unpaused state.
Logs & Output