From b3904182c68401f9573dadd09035bb8cd87a96a8 Mon Sep 17 00:00:00 2001
From: Paul Yang <github@ardorful.com>
Date: Fri, 4 Aug 2023 13:34:53 -0700
Subject: [PATCH] Add Github workflow to populate the persistent source schema.

---
 ...ine-populate-persistent-source-schema.yaml | 95 +++++++++++++++++++
 GLOSSARY.md                                   | 13 +++
 Makefile                                      | 23 +++++
 3 files changed, 131 insertions(+)
 create mode 100644 .github/workflows/cd-sql-engine-populate-persistent-source-schema.yaml
 create mode 100644 GLOSSARY.md

diff --git a/.github/workflows/cd-sql-engine-populate-persistent-source-schema.yaml b/.github/workflows/cd-sql-engine-populate-persistent-source-schema.yaml
new file mode 100644
index 0000000000..cd9f3ff579
--- /dev/null
+++ b/.github/workflows/cd-sql-engine-populate-persistent-source-schema.yaml
@@ -0,0 +1,95 @@
+# See [Persistent Source Schema](/GLOSSARY.md#persistent-source-schema)
+# Populating the source schema via this workflow ensures that it's done with the same settings as the tests.
+
+name: Reload Test Data in SQL Engines
+
+# We don't want multiple workflows trying to create the same table.
+concurrency:
+  group: POPULATE_PERSISTENT_SOURCE_SCHEMA
+  cancel-in-progress: true
+
+on:
+  pull_request:
+    types: [labeled, synchronize]
+  workflow_dispatch:
+
+env:
+  PYTHON_VERSION: "3.8"
+  ADDITIONAL_PYTEST_OPTIONS: "--use-persistent-source-schema"
+
+jobs:
+  snowflake-populate:
+    environment: DW_INTEGRATION_TESTS
+    if: ${{ github.event.action != 'labeled' || github.event.label.name == 'Reload Test Data in SQL Engines' }}
+    name: Snowflake
+    runs-on: ubuntu-latest
+    steps:
+      - name: Check-out the repo
+        uses: actions/checkout@v3
+
+      - name: Populate w/Python ${{ env.PYTHON_VERSION }}
+        uses: ./.github/actions/run-mf-tests
+        with:
+          python-version: ${{ env.PYTHON_VERSION }}
+          mf_sql_engine_url: ${{ secrets.MF_SNOWFLAKE_URL }}
+          mf_sql_engine_password: ${{ secrets.MF_SNOWFLAKE_PWD }}
+          parallelism: 1
+          additional-pytest-options: ${{ env.ADDITIONAL_PYTEST_OPTIONS }}
+          make-target: "populate-persistent-source-schema-snowflake"
+
+  redshift-populate:
+    environment: DW_INTEGRATION_TESTS
+    name: Redshift
+    if: ${{ github.event.action != 'labeled' || github.event.label.name == 'Reload Test Data in SQL Engines' }}
+    runs-on: ubuntu-latest
+    steps:
+      - name: Check-out the repo
+        uses: actions/checkout@v3
+
+      - name: Populate w/Python ${{ env.PYTHON_VERSION }}
+        uses: ./.github/actions/run-mf-tests
+        with:
+          python-version: ${{ env.PYTHON_VERSION }}
+          mf_sql_engine_url: ${{ secrets.MF_REDSHIFT_URL }}
+          mf_sql_engine_password: ${{ secrets.MF_REDSHIFT_PWD }}
+          parallelism: 1
+          additional-pytest-options: ${{ env.ADDITIONAL_PYTEST_OPTIONS }}
+          make-target: "populate-persistent-source-schema-redshift"
+
+  bigquery-populate:
+    environment: DW_INTEGRATION_TESTS
+    name: BigQuery
+    if: ${{ github.event.action != 'labeled' || github.event.label.name == 'Reload Test Data in SQL Engines' }}
+    runs-on: ubuntu-latest
+    steps:
+      - name: Check-out the repo
+        uses: actions/checkout@v3
+
+      - name: Populate w/Python ${{ env.PYTHON_VERSION }}
+        uses: ./.github/actions/run-mf-tests
+        with:
+          python-version: ${{ env.PYTHON_VERSION }}
+          MF_SQL_ENGINE_URL: ${{ secrets.MF_BIGQUERY_URL }}
+          MF_SQL_ENGINE_PASSWORD: ${{ secrets.MF_BIGQUERY_PWD }}
+          parallelism: 1
+          additional-pytest-options: ${{ env.ADDITIONAL_PYTEST_OPTIONS }}
+          make-target: "populate-persistent-source-schema-bigquery"
+
+  databricks-populate:
+    environment: DW_INTEGRATION_TESTS
+    name: Databricks SQL Warehouse
+    if: ${{ github.event.action != 'labeled' || github.event.label.name == 'Reload Test Data in SQL Engines' }}
+    runs-on: ubuntu-latest
+    steps:
+      - name: Check-out the repo
+        uses: actions/checkout@v3
+
+      - name: Populate w/Python ${{ env.PYTHON_VERSION }}
+        uses: ./.github/actions/run-mf-tests
+        with:
+          python-version: ${{ env.PYTHON_VERSION }}
+          mf_sql_engine_url: ${{ secrets.MF_DATABRICKS_SQL_WAREHOUSE_URL }}
+          mf_sql_engine_password: ${{ secrets.MF_DATABRICKS_PWD }}
+          parallelism: 1
+          additional-pytest-options: ${{ env.ADDITIONAL_PYTEST_OPTIONS }}
+          make-target: "populate-persistent-source-schema-databricks"
diff --git a/GLOSSARY.md b/GLOSSARY.md
new file mode 100644
index 0000000000..ebb047bef8
--- /dev/null
+++ b/GLOSSARY.md
@@ -0,0 +1,13 @@
+# Glossary
+
+## Persistent source schema
+Many tests generate and execute SQL that depend on tables containing test data. By default, a
+pytest fixture creates a temporary schema and populates it with the tables that are required by
+the tests. This schema is referred to the source schema. Creating the source schema (and
+the associated tables) can be a slow process for some SQL engines. Since these tables generally
+do not change often, functionality was added to use a source schema that is assumed to already
+exist when running tests and persists between runs (a persistent source schema). In addition,
+functionality was added to create the persistent source schema based on table definitions in the
+repo. Because the name of the source schema is generated based on the hash of the data that's
+supposed to be in the schema, the creating and populating the persistent source schema should
+not be done concurrently as there are race conditions when creating tables and inserting data.
diff --git a/Makefile b/Makefile
index bc2d10b863..1158c61c24 100644
--- a/Makefile
+++ b/Makefile
@@ -7,6 +7,9 @@ PARALLELISM = "auto"
 # Additional command line options to pass to pytest.
 ADDITIONAL_PYTEST_OPTIONS = ""
 
+# Pytest that can populate the persistent source schema
+POPULATE_PERSISTENT_SOURCE_SCHEMA = metricflow/test/source_schema_tools.py::populate_source_schema
+
 # Install Hatch package / project manager
 .PHONY: install-hatch
 install-hatch:
@@ -23,22 +26,42 @@ test-postgresql:
 
 # Engine-specific test environments. In most cases you should run these with
 # `make -e ADDITIONAL_PYTEST_OPTIONS="--use-persistent-source-schema" test-<engine_type>`
+#
+# See: [Persistent Source Schema](/GLOSSARY.md#persistent-source-schema)
 .PHONY: test-bigquery
 test-bigquery:
 	hatch -v run bigquery-env:pytest -vv -n $(PARALLELISM) $(ADDITIONAL_PYTEST_OPTIONS) metricflow/test/
 
+.PHONY: populate-persistent-source-schema-bigquery
+populate-persistent-source-schema-bigquery:
+	hatch -v run bigquery-env:pytest -vv $(ADDITIONAL_PYTEST_OPTIONS) $(POPULATE_PERSISTENT_SOURCE_SCHEMA)
+
 .PHONY: test-databricks
 test-databricks:
 	hatch -v run databricks-env:pytest -vv -n $(PARALLELISM) $(ADDITIONAL_PYTEST_OPTIONS) metricflow/test/
 
+.PHONY: populate-persistent-source-schema-databricks
+populate-persistent-source-schema-databricks:
+	hatch -v run databricks-env:pytest -vv $(ADDITIONAL_PYTEST_OPTIONS) $(POPULATE_PERSISTENT_SOURCE_SCHEMA)
+
 .PHONY: test-redshift
 test-redshift:
 	hatch -v run redshift-env:pytest -vv -n $(PARALLELISM) $(ADDITIONAL_PYTEST_OPTIONS) metricflow/test/
 
+.PHONY: populate-persistent-source-schema-redshift
+populate-persistent-source-schema-redshift:
+	hatch -v run redshift-env:pytest -vv $(ADDITIONAL_PYTEST_OPTIONS) $(POPULATE_PERSISTENT_SOURCE_SCHEMA)
+
+
 .PHONY: test-snowflake
 test-snowflake:
 	hatch -v run snowflake-env:pytest -vv -n $(PARALLELISM) $(ADDITIONAL_PYTEST_OPTIONS) metricflow/test/
 
+.PHONY: populate-persistent-source-schema-snowflake
+populate-persistent-source-schema-snowflake:
+	hatch -v run snowflake-env:pytest -vv $(ADDITIONAL_PYTEST_OPTIONS) $(POPULATE_PERSISTENT_SOURCE_SCHEMA)
+
+
 .PHONY: lint
 lint:
 	hatch -v run dev-env:pre-commit run --all-files