WIP on CI

.github/actions/test-data-plane/action.yaml

@@ -0,0 +1,57 @@
+name: 'Test Data Plane'
+description: 'Runs tests on the Pinecone data plane'
+
+# - pod vs serverless
+# - grpc vs rest
+# - metric -> vector vs sparse vector
+# - namespace: default vs custom
+# - environment: free vs paid
+# - with metadata vs without metadata
+
+inputs:
+  metric:
+    description: 'The metric of the index'
+    required: true
+    default: 'cosine'
+  spec:
+    description: 'The deploy spec of the index'
+    required: false
+    default: ''
+  use_grpc:
+    description: 'Whether to use gRPC or REST'
+    required: false
+    default: 'true'
+  freshness_sleep_seconds:
+    description: 'The number of seconds to wait for the index to become fresh'
+    required: false
+    default: '60'
+
+outputs:
+  index_name:
+    description: 'The name of the index, including randomized suffix'
+    value: ${{ steps.create-index.outputs.index_name }}
+
+runs:
+  using: 'composite'
+  steps:
+    - name: Set up Python
+      uses: actions/setup-python@v4
+      with:
+        python-version: 3.9
+
+    - name: Setup Poetry
+      uses: ./.github/actions/setup-poetry
+      with:
+        include_grpc: ${{ inputs.use_grpc }}
+        include_dev: 'true'
+
+    - name: Run data plane tests
+      id: data-plane-tests
+      shell: bash
+      run: poetry run pytest tests/integration/data
+      env:
+        PINECONE_API_KEY: ${{ inputs.PINECONE_API_KEY }}
+        USE_GRPC: ${{ inputs.use_grpc }}
+        METRIC: ${{ inputs.metric }}
+        SPEC: ${{ inputs.spec }}
+        FRESHNESS_SLEEP_SECONDS: ${{ inputs.freshness_sleep_seconds }}

.github/workflows/testing-grpc.yaml

@@ -34,6 +34,27 @@ jobs:
           # so we don't add new regressions.
           poetry run mypy pinecone --exclude pinecone/core
 
+  integration-grpc-data-plane:
+    name: Integration tests (GRPC, data plane)
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        metric:
+          - cosine
+          - dotproduct
+        spec:
+          # These need to be json strings
+          - '{ "serverless": { "region": "us-west-2", "cloud": "aws" }}'
+          - '{ "pod": { "environment": "us-east1-gcp" }}'
+
+    steps:
+      - uses: actions/checkout@v4
+      - uses: ./.github/actions/test-data-plane
+        with:
+          include_grpc: true
+          metric: ${{ matrix.metric }}
+          spec: ${{ matrix.spec }}
+
   dependency-matrix-grpc-setup:
     name: Deps setup (GRPC)
     runs-on: ubuntu-latest

.github/workflows/testing-rest.yaml

@@ -83,6 +83,26 @@ jobs:
           SERVERLESS_CLOUD: ${{ matrix.testConfig.serverless.cloud }}
           SERVERLESS_REGION: ${{ matrix.testConfig.serverless.region }}
 
+  integration-rest-data-plane:
+    name: Integration tests (GRPC, data plane)
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        metric:
+          - cosine
+          - dotproduct
+        spec:
+          # These need to be json strings
+          - '{ "serverless": { "region": "us-west-2", "cloud": "aws" }}'
+          - '{ "pod": { "environment": "us-east1-gcp" }}'
+    steps:
+        - uses: actions/checkout@v4
+        - uses: ./.github/actions/test-data-plane
+          with:
+            include_grpc: false
+            metric: ${{ matrix.metric }}
+            spec: ${{ matrix.spec }}
+
   dependency-matrix-rest-setup:
     name: Deps setup (REST)
     runs-on: ubuntu-latest

tests/integration/conftest.py → tests/integration/control/conftest.py

@@ -3,7 +3,7 @@
 import random
 import time
 from pinecone import Pinecone, NotFoundException, PineconeApiException
-from .helpers.helpers import generate_index_name, get_environment_var
+from ..helpers import generate_index_name, get_environment_var
 
 @pytest.fixture()
 def client():

tests/integration/data/conftest.py

@@ -0,0 +1,51 @@
+import pytest
+import os
+import json
+from ..helpers import get_environment_var, random_string
+
+# Test matrix needs to consider the following dimensions:
+# - pod vs serverless
+# - grpc vs rest
+# - metric -> vector vs sparse vector
+# - namespace: default vs custom
+# - environment: free vs paid
+# - with metadata vs without metadata
+
+@pytest.fixture
+def api_key():
+    return get_environment_var('PINECONE_API_KEY')
+
+@pytest.fixture
+def client(api_key):
+    use_grpc = os.environ.get('USE_GRPC', 'false') == 'true'
+    if use_grpc:
+        from pinecone.grpc import PineconeGRPC
+        return PineconeGRPC(api_key=api_key)
+    else:
+        from pinecone import Pinecone
+        return Pinecone(api_key=api_key)
+
+@pytest.fixture
+def metric():
+    return get_environment_var('METRIC', 'cosine')
+
+@pytest.fixture
+def spec():
+    return json.loads(get_environment_var('SPEC'))
+
+@pytest.fixture
+def index_name():
+    return 'dataplane-' + random_string(20)
+
+@pytest.fixture
+def namespace():
+    return random_string(10)
+
+@pytest.fixture
+def index_host(client, index_name, metric, spec):
+    client.create_index(name=index_name, dimension=2, metric=metric, spec=spec)
+    description = client.describe_index(name=index_name)
+    return description.host
+
+def sleep_t():
+    return int(os.environ.get('FRESHNESS_SLEEP_SECONDS', 60))

tests/integration/data/cosine/test_upsert.py

@@ -0,0 +1,54 @@
+import pytest
+import time
+from pinecone import Vector
+
+def test_upsert_to_default_namespace(client, index_name, sleep_t):
+    expected_dimension = 2
+    desc = client.describe_index(index_name)
+    assert desc.dimension == expected_dimension
+    assert desc.metric == 'cosine'
+
+    idx = client.Index(index_name)
+
+    # Upsert with tuples
+    idx.upsert(vectors=[
+        ('1', [1.0, 2.0]), 
+        ('2', [3.0, 4.0]),
+        ('3', [5.0, 6.0])
+    ])
+
+    # Upsert with objects
+    idx.upsert(vectors=[
+        Vector('4', [7.0, 8.0]),
+        Vector('5', [9.0, 10.0]),
+        Vector('6', [11.0, 12.0])
+    ])
+
+    # Upsert with dict
+    idx.upsert(vectors=[
+        {'id': '7', 'values': [13.0, 14.0]},
+        {'id': '8', 'values': [15.0, 16.0]},
+        {'id': '9', 'values': [17.0, 18.0]}
+    ])
+
+    time.sleep(sleep_t)
+
+    # Check the vector count reflects some data has been upserted
+    stats = idx.describe_index_stats()
+    assert stats.vector_count == 9
+
+
+def test_upsert_to_custom_namespace(client, index_name, namespace):
+    expected_dimension = 2
+    assert client.describe_index(index_name).dimension == expected_dimension
+
+    idx = client.Index(index_name)
+
+    # Upsert with tuples
+    idx.upsert(vectors=[
+        ('1', [1.0, 2.0]), 
+        ('2', [3.0, 4.0]),
+        ('3', [5.0, 6.0])
+        ], 
+        namespace=namespace
+    )