Google Cloud Platform - Billing, how do I find the name of the instance of a product that is costing money?

Question

I go into billing and filter by project & group by SKU. This gives me the individual services that I'm being charged for (view below). I can see that Compute Engine is costing me £5.59 per month. If I go into Compute Engine I don't have any instances. In this particular example I discovered that had a Serverless VPC Connector running and I think that is using the Service but it is just a guess.

How do I find out what specific product instances that I have enabled are costing me money? How, in this example, would I find out from Billing that it is the Serverless VPC Connector that is costing me money and linked to that Compute Engine instance?

(I've had other situations where I'm being billed for an IP address but not the name of that IP address or what products / instances of products are consuming the IP address).

Answer 1

Interesting...

It looks like indeed looking at https://cloud.google.com/vpc/pricing#serverless-vpc-pricing serverless VPC connectors could be billed as e2-micro

---

While in many cases you will see billing report actually try to associate your billing to human understandable service like so (under service and/or SKU columns) it doesn't seem to be always the case.

---

One way to try to resolve that I guess would be to follow this path

https://cloud.google.com/blog/topics/cost-management/use-labels-to-gain-visibility-into-gcp-resource-usage-and-spending

Following instructions and once you get your billing exported to big query you could run the following:

select
sku.id,
l.key

from
YOUR_PROJECT.YOUR_BILLING_EXPORT_DATASET.YOUR_BILLING_TABLE
,UNNEST(labels) as l

where
 sku.id = '9FE0-8F60-A9FO' # your SKU id in question

and labels should give you a better idea of whats what.

---

This should give you an overall direction, perhaps there is a way to parse this out without exporting to BQ (first) and querying it (second) as this may also cost you money.

---

PS. Labels are always a good idea when you create services/resources explicitly.