Cloud SQL Admin API - Sudden and ongoing spike

[Hudson Kanizawa]

Since last monday (15/10) a notice a sudden spike on Cloud SQL Admin API's queries, but we didn't make any changes in our application or in our environment recently. 

Is there any place where I can see which queries are these?

I'm looking everywhere but I can't find the origin of this spike.

[Katayoon (Cloud Platform Support)]

Hi Hudson,

I should note that Cloud SQL Admin API is for administering your instances programmatically, and you can make Advanced Logs Filters to get more information on the request sent to this API. As you can find in Cloud SQL Admin API console, the service name is "sqladmin.googleapis.com". So you may use this service name as a filter in your Logs Viewer. 

[Hudson Kanizawa]

Thank you Katayoon,

I'm trying to filter logs, but I'm just finding register from [ protoPayload.serviceName= "cloudsql.googleapis.com"].

Filtering by [ protoPayload.serviceName= "sqladmin.googleapis.com"] it gives me no logs, even on days showing those spike on charts.

[Olu]

The serviceName, cloudsql.googleapis.com filters logs associated with the CloudSQL Instance, rather than the CloudSQL admin API. More specifically, the service name for the Cloud SQL Admin API is "sqladmin.googleapis.com". 

If no results are returned from using your Advanced Filter with serviceName= "sqladmin.googleapis.com", I recommend reporting this issue about the spike on Cloud SQL Admin API's queries by creating an issue on the Issue tracker. An engineer should be able to work with you reviewing the details of your Stackdriver logs to understand the Spikes in your usage of the API.