Integration test reorg

.github/actions/index-create/action.yml

@@ -17,11 +17,15 @@ inputs:
   dimension:
     description: 'The dimension of the index'
     required: false
-    default: '3'
+    default: ''
   metric:
     description: 'The metric of the index'
     required: false
     default: 'cosine'
+  vector_type:
+    description: 'The type of the index'
+    required: false
+    default: 'dense'
   PINECONE_API_KEY:
     description: 'The Pinecone API key'
     required: true
@@ -36,6 +40,10 @@ outputs:
     description: 'The name of the index, including randomized suffix'
     value: ${{ steps.create-index.outputs.index_name }}
 
+  index_host:
+    description: 'The host of the index'
+    value: ${{ steps.create-index.outputs.index_host }}
+
 runs:
   using: 'composite'
   steps:
@@ -52,5 +60,6 @@ runs:
         NAME_PREFIX: ${{ inputs.name_prefix }}
         REGION: ${{ inputs.region }}
         CLOUD: ${{ inputs.cloud }}
+        VECTOR_TYPE: ${{ inputs.vector_type }}
         DIMENSION: ${{ inputs.dimension }}
         METRIC: ${{ inputs.metric }}

.github/actions/index-create/create.py

@@ -1,9 +1,9 @@
 import os
-import re
 import random
 import string
-from datetime import datetime
+import uuid
 from pinecone import Pinecone
+from datetime import datetime
 
 
 def read_env_var(name):
@@ -22,39 +22,9 @@ def write_gh_output(name, value):
         print(f"{name}={value}", file=fh)
 
 
-def generate_index_name(test_name: str) -> str:
-    github_actor = os.getenv("GITHUB_ACTOR", None)
-    user = os.getenv("USER", None)
-    index_owner = github_actor or user
-
-    formatted_date = datetime.now().strftime("%Y%m%d-%H%M%S%f")[:-3]
-
-    github_job = os.getenv("GITHUB_JOB", None)
-
-    if test_name.startswith("test_"):
-        test_name = test_name[5:]
-
-    # Remove trailing underscore, if any
-    if test_name.endswith("_"):
-        test_name = test_name[:-1]
-
-    name_parts = [index_owner, formatted_date, github_job, test_name]
-    index_name = "-".join([x for x in name_parts if x is not None])
-
-    # Remove invalid characters
-    replace_with_hyphen = re.compile(r"[\[\(_,\s]")
-    index_name = re.sub(replace_with_hyphen, "-", index_name)
-    replace_with_empty = re.compile(r"[\]\)\.]")
-    index_name = re.sub(replace_with_empty, "", index_name)
-
-    max_length = 45
-    index_name = index_name[:max_length]
-
-    # Trim final character if it is not alphanumeric
-    if index_name.endswith("_") or index_name.endswith("-"):
-        index_name = index_name[:-1]
-
-    return index_name.lower()
+def generate_index_name(name_prefix: str) -> str:
+    name = name_prefix.lower() + "-" + str(uuid.uuid4())
+    return name[:45]
 
 
 def get_tags():
@@ -74,15 +44,35 @@ def get_tags():
 
 def main():
     pc = Pinecone(api_key=read_env_var("PINECONE_API_KEY"))
-    index_name = generate_index_name(read_env_var("NAME_PREFIX") + random_string(20))
+    index_name = generate_index_name(read_env_var("NAME_PREFIX"))
+    dimension_var = read_env_var("DIMENSION")
+    if dimension_var is not None and dimension_var != "":
+        dimension = int(dimension_var)
+    else:
+        dimension = None
+
+    vector_type_var = read_env_var("VECTOR_TYPE")
+    if vector_type_var is not None and vector_type_var != "":
+        vector_type = vector_type_var
+    else:
+        vector_type = None
+
+    metric = read_env_var("METRIC")
+    cloud = read_env_var("CLOUD")
+    region = read_env_var("REGION")
+    tags = get_tags()
+
     pc.create_index(
         name=index_name,
-        metric=read_env_var("METRIC"),
-        dimension=int(read_env_var("DIMENSION")),
-        spec={"serverless": {"cloud": read_env_var("CLOUD"), "region": read_env_var("REGION")}},
-        tags=get_tags(),
+        metric=metric,
+        dimension=dimension,
+        vector_type=vector_type,
+        tags=tags,
+        spec={"serverless": {"cloud": cloud, "region": region}},
     )
+    description = pc.describe_index(name=index_name)
     write_gh_output("index_name", index_name)
+    write_gh_output("index_host", description.host)
 
 
 if __name__ == "__main__":

.github/actions/run-integration-test/action.yaml

@@ -14,11 +14,29 @@ inputs:
   PINECONE_ADDITIONAL_HEADERS:
     description: 'Additional headers to send with the request'
     required: false
-    default: '{"sdk-test-suite": "pinecone-python-client"}'
+    default: '{"sdk-test-suite": "pinecone-python-client", "x-environment": "preprod-aws-0"}'
   use_grpc:
     description: 'Whether to use gRPC or REST'
     required: false
     default: 'false'
+  PINECONE_CLIENT_ID:
+    description: 'The client ID to use for admin tests'
+    required: false
+  PINECONE_CLIENT_SECRET:
+    description: 'The client secret to use for admin tests'
+    required: false
+  INDEX_HOST_DENSE:
+    description: 'The host of the dense index for db data tests'
+    required: false
+  INDEX_HOST_SPARSE:
+    description: 'The host of the sparse index for db data tests'
+    required: false
+  pytest_splits:
+    description: 'Number of shards to split tests into (for test sharding)'
+    required: false
+  pytest_group:
+    description: 'Which shard to run (1-indexed, for test sharding)'
+    required: false
 
 runs:
   using: 'composite'
@@ -33,9 +51,23 @@ runs:
     - name: Run tests
       id: run-tests
       shell: bash
-      run: poetry run pytest tests/integration/${{ inputs.test_suite }} --retries 2 --retry-delay 35 -s -vv --log-cli-level=DEBUG --durations=20
+      run: |
+        PYTEST_ARGS=""
+        if [ -n "${{ inputs.pytest_splits }}" ] && [ -n "${{ inputs.pytest_group }}" ]; then
+          PYTEST_ARGS="--splits=${{ inputs.pytest_splits }} --group=${{ inputs.pytest_group }}"
+        fi
+        poetry run pytest ${{ inputs.test_suite }} \
+          $PYTEST_ARGS \
+          --retries 2 \
+          --retry-delay 35 \
+          --log-cli-level=DEBUG \
+          --durations=25 \
+          -s -vv
       env:
         PINECONE_API_KEY: ${{ steps.decrypt-api-key.outputs.decrypted_secret }}
         PINECONE_ADDITIONAL_HEADERS: ${{ inputs.PINECONE_ADDITIONAL_HEADERS }}
+        PINECONE_CLIENT_ID: ${{ inputs.PINECONE_CLIENT_ID }}
+        PINECONE_CLIENT_SECRET: ${{ inputs.PINECONE_CLIENT_SECRET }}
         USE_GRPC: ${{ inputs.use_grpc }}
-        SKIP_WEIRD: 'true'
+        INDEX_HOST_DENSE: ${{ inputs.INDEX_HOST_DENSE }}
+        INDEX_HOST_SPARSE: ${{ inputs.INDEX_HOST_SPARSE }}