Vector Search Index Serving e2-standard-16 in europe-west1 Costs

Google Cloud Build with AI Custom ML & MLOps
1

Hi,

I have been experimenting with RAG by creating a VectorSearchVectorStore and MatchingEngineIndexEndpoint. I noticed that the cost is high, so I deleted the index and endpoint and switched to using ChromaDb. However, I am still being billed for Vector Search Index Serving e2-standard-16 in europe-west1 despite having deleted the index and endpoint:

ahmed94_0-1733905298871.png
ahmed94_0-1733905298871.png823×204 14.6 KB

My code:

index = aiplatform.MatchingEngineIndex.create_tree_ah_index(
display_name=“langchain_index2¡”,
dimensions=768, # Dimensiones de los embeddings de texto
approximate_neighbors_count=200,
leaf_node_embedding_count=600,
leaf_nodes_to_search_percent=7,
description=“Text-based LangChain Index”
)

index_endpoint = aiplatform.MatchingEngineIndexEndpoint.create(
display_name=“langchain_index_endpoint”,
description=“Text-based LangChain Index Endpoint”,
public_endpoint_enabled=True,
)
index_endpoint.deploy_index(
index=index,
deployed_index_id=“langchain_deployed_index”
)

embeddings = VertexAIEmbeddings(model_name=“textembedding-gecko@003”)

aiplatform.init(project=PROJECT_ID, location=LOCATION)

index_endpoint = aiplatform.MatchingEngineIndexEndpoint(

index_endpoint_name=ENDPOINT_NAME

)

vectorstore = VectorSearchVectorStore.from_components(
project_id=PROJECT_ID,
region=LOCATION,
gcs_bucket_name=GCS_BUCKET,
index_id=index_id,
endpoint_id=endpoint_id,
embedding=embeddings,
stream_update=False
)

Delete endpoint and index

vs_endpoint.undeploy_all()
vs_endpoint.delete()

vs_index.delete()

###########

Now for retriever I am using chroma db

db = Chroma(persist_directory=LOCAL_PERSIST_PATH, embedding_function=embeddings)
retriever = db.as_retriever(search_type=“similarity”, search_kwargs={“k”: 3})

Can you help me please, I don’t know what to do

2

Hi @ahmed94,

Welcome to Google Cloud Community!

It seems like you’ve deleted the index and endpoint, yet charges persist for the e2-standard-16 instance in Europe-west1. This might be due to residual resources or configurations.

Here are potential ways that might help with your use case:

  • Confirm the deletion: Double-check that the index and endpoint have been successfully deleted. Sometimes, there may be a delay in the billing system reflecting these changes.
  • Review your remaining resources: Make sure that there are no residual resources or instances still active or being billed. To verify it, you may use Google Cloud Console.
  • Check your billing details: Check your billing details to identify any other resources or services using the e2-standard-16 instance. Unexpected charges may arise from other services or misconfigurations.
  • Review your billing reports: Review your billing reports to identify any other resources or instances that may still be active and potentially leading to charges.

If you need help with your Cloud Billing account, Google Cloud offers free Cloud Billing Support to all customers. If you’d like to talk to a support representative about a Cloud Billing question or a billing-related issue, visit our Cloud Billing Support page for contact options.

Was this helpful? If so, please accept this answer as “Solution”. If you need additional assistance, reply here within 2 business days and I’ll be happy to help.