Detailed Info on Batch Ingestion for Vertex AI Feature Store

[subha_m]

I'm curious about the costs associated with batch ingestion within Google Vertex AI Feature Store. Can someone break down how the pricing is determined for this specific service?
I could see the streaming ingestion price details but I couldn't see the batch ingestion price details in the link https://cloud.google.com/vertex-ai/pricing
Could anyone kindly assist me?

[kvandres · Accepted solution]

Hi @subha_m,

Based in the documentation: https://cloud.google.com/vertex-ai/pricing

A node hour represents the time a virtual machine spends serving feature data or waiting in a ready state to handle feature data requests.

---

Since it doesn't specifically include the batch ingestion pricing, you will be charged for the nodes that are in ready state while they are waiting to handle requests, for example, since you can't scale down the node to zero, you will be charged for the 1 node that is in ready state while it is waiting to handle the request.

[subha_m]

Thanks for the reply @kvandres .

Apart from node cost, Is there any specific cost involved in displaying features in feature store?