Skip to content

Commit

Permalink
feat(vertexai): Support google_vertex_ai_feature_group (#9520) (#627)
Browse files Browse the repository at this point in the history
[upstream:49252d7a95dc28f660ad9be07c260f02e23d51fb]

Signed-off-by: Modular Magician <[email protected]>
  • Loading branch information
modular-magician authored Dec 19, 2023
1 parent fd22a15 commit 7de7afd
Show file tree
Hide file tree
Showing 4 changed files with 144 additions and 0 deletions.
15 changes: 15 additions & 0 deletions vertex_ai_feature_group/backing_file.tf
Original file line number Diff line number Diff line change
@@ -0,0 +1,15 @@
# This file has some scaffolding to make sure that names are unique and that
# a region and zone are selected when you try to create your Terraform resources.

locals {
name_suffix = "${random_pet.suffix.id}"
}

resource "random_pet" "suffix" {
length = 2
}

provider "google" {
region = "us-central1"
zone = "us-central1-c"
}
43 changes: 43 additions & 0 deletions vertex_ai_feature_group/main.tf
Original file line number Diff line number Diff line change
@@ -0,0 +1,43 @@
resource "google_vertex_ai_feature_group" "feature_group" {
name = "example_feature_group-${local.name_suffix}"
description = "A sample feature group"
region = "us-central1"
labels = {
label-one = "value-one"
}
big_query {
big_query_source {
# The source table must have a column named 'feature_timestamp' of type TIMESTAMP.
input_uri = "bq://${google_bigquery_table.sample_table.project}.${google_bigquery_table.sample_table.dataset_id}.${google_bigquery_table.sample_table.table_id}"
}
entity_id_columns = ["feature_id"]
}
}

resource "google_bigquery_dataset" "sample_dataset" {
dataset_id = "job_load-${local.name_suffix}_dataset"
friendly_name = "test"
description = "This is a test description"
location = "US"
}

resource "google_bigquery_table" "sample_table" {
deletion_protection = false
dataset_id = google_bigquery_dataset.sample_dataset.dataset_id
table_id = "job_load-${local.name_suffix}_table"

schema = <<EOF
[
{
"name": "feature_id",
"type": "STRING",
"mode": "NULLABLE"
},
{
"name": "feature_timestamp",
"type": "TIMESTAMP",
"mode": "NULLABLE"
}
]
EOF
}
7 changes: 7 additions & 0 deletions vertex_ai_feature_group/motd
Original file line number Diff line number Diff line change
@@ -0,0 +1,7 @@
===

These examples use real resources that will be billed to the
Google Cloud Platform project you use - so make sure that you
run "terraform destroy" before quitting!

===
79 changes: 79 additions & 0 deletions vertex_ai_feature_group/tutorial.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,79 @@
# Vertex Ai Feature Group - Terraform

## Setup

<walkthrough-author name="[email protected]" analyticsId="UA-125550242-1" tutorialName="vertex_ai_feature_group" repositoryUrl="https://github.com/terraform-google-modules/docs-examples"></walkthrough-author>

Welcome to Terraform in Google Cloud Shell! We need you to let us know what project you'd like to use with Terraform.

<walkthrough-project-billing-setup></walkthrough-project-billing-setup>

Terraform provisions real GCP resources, so anything you create in this session will be billed against this project.

## Terraforming!

Let's use {{project-id}} with Terraform! Click the Cloud Shell icon below to copy the command
to your shell, and then run it from the shell by pressing Enter/Return. Terraform will pick up
the project name from the environment variable.

```bash
export GOOGLE_CLOUD_PROJECT={{project-id}}
```

After that, let's get Terraform started. Run the following to pull in the providers.

```bash
terraform init
```

With the providers downloaded and a project set, you're ready to use Terraform. Go ahead!

```bash
terraform apply
```

Terraform will show you what it plans to do, and prompt you to accept. Type "yes" to accept the plan.

```bash
yes
```


## Post-Apply

### Editing your config

Now you've provisioned your resources in GCP! If you run a "plan", you should see no changes needed.

```bash
terraform plan
```

So let's make a change! Try editing a number, or appending a value to the name in the editor. Then,
run a 'plan' again.

```bash
terraform plan
```

Afterwards you can run an apply, which implicitly does a plan and shows you the intended changes
at the 'yes' prompt.

```bash
terraform apply
```

```bash
yes
```

## Cleanup

Run the following to remove the resources Terraform provisioned:

```bash
terraform destroy
```
```bash
yes
```

0 comments on commit 7de7afd

Please sign in to comment.