[Chore] Automatically cleanup old resources each night (#400)

## Problem

Sometimes when test cleanup steps fail, indexes and collections get left
behind.

## Solution

Create a nightly job to cleanup leftover indexes. Inspect the names of
each index and collection to see whether they are more than 24 hours old
prior to deleting. This should prevent deleting resources out from
underneath any tests that may be running at the same time as the delete
job.

## Type of Change

- [x] Infrastructure change (CI configs, etc)

Author: jhamon

.github/workflows/cleanup-nightly.yaml

@@ -0,0 +1,17 @@
+name: 'Cleanup All Indexes/Collections (Nightly)'
+
+on:
+    schedule:
+      - cron:  '5 22 * * *' # 5 minutes after 10pm UTC, every day
+
+jobs:
+  cleanup-all:
+    name: Cleanupu all indexes/collections
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - name: Cleanup all
+        uses: ./.github/actions/cleanup-all
+        with:
+          PINECONE_API_KEY: ${{ secrets.PINECONE_API_KEY }}
+          DELETE_ALL: false

.github/workflows/cleanup.yaml

@@ -12,4 +12,5 @@ jobs:
       - name: Cleanup all
         uses: ./.github/actions/cleanup-all
         with:
-          PINECONE_API_KEY: ${{ secrets.PINECONE_API_KEY }}
+          PINECONE_API_KEY: ${{ secrets.PINECONE_API_KEY }}
+          DELETE_ALL: true

scripts/cleanup-all.py

@@ -1,10 +1,10 @@
 import os
+import re
 from pinecone import Pinecone
+from datetime import datetime, timedelta
 
 
-def main():
-    pc = Pinecone(api_key=os.environ.get("PINECONE_API_KEY", None))
-
+def delete_everything(pc):
     for collection in pc.list_collections().names():
         try:
             print("Deleting collection: " + collection)
@@ -22,5 +22,65 @@ def main():
             pass
 
 
+def parse_date(resource_name):
+    match = re.search(r"-\d{8}-", resource_name)
+    if match:
+        date_string = match.group(0).strip("-")
+        return datetime.strptime(date_string, "%Y%m%d")
+    else:
+        return None
+
+
+def is_resource_old(resource_name):
+    print(f"Checking resource name: {resource_name}")
+    resource_datetime = parse_date(resource_name)
+    if resource_datetime is None:
+        return False
+    current_time = datetime.now()
+
+    # Calculate the difference
+    time_difference = current_time - resource_datetime
+
+    # Check if the time difference is greater than 24 hours
+    print(f"Resource timestamp: {resource_datetime}")
+    print(f"Time difference: {time_difference}")
+    return time_difference > timedelta(hours=24)
+
+
+def delete_old(pc):
+    for collection in pc.list_collections().names():
+        if is_resource_old(collection):
+            try:
+                print("Deleting collection: " + collection)
+                pc.delete_collection(collection)
+            except Exception as e:
+                print("Failed to delete collection: " + collection + " " + str(e))
+                pass
+        else:
+            print("Skipping collection, not old enough: " + collection)
+
+    for index in pc.list_indexes().names():
+        if is_resource_old(index):
+            try:
+                print("Deleting index: " + index)
+                pc.delete_index(index)
+            except Exception as e:
+                print("Failed to delete index: " + index + " " + str(e))
+                pass
+        else:
+            print("Skipping index, not old enough: " + index)
+
+
+def main():
+    pc = Pinecone(api_key=os.environ.get("PINECONE_API_KEY", None))
+
+    if os.environ.get("DELETE_ALL", None) == "true":
+        print("Deleting everything")
+        delete_everything(pc)
+    else:
+        print("Deleting old resources")
+        delete_old(pc)
+
+
 if __name__ == "__main__":
     main()