You need to enable or disable AutoSuspend to handle resources when the auto-scaler has scaled back to the last executor group. You can control the time that the original warehouse executor group idles after all other groups scale down and release their executors. The JDBC endpoint lives on to respond to queries from the result cache or statistics, but expensive executors no longer run.
When the auto-scaler increases or decreases the number of executor groups in the Virtual Warehouse, it can take several minutes before the scaling up or down takes effect. This slight delay is caused by the time required by your cloud provider to provision clusters.
When no queries are sent to an executor group, resources scale down and executors are released. When all executor groups are scaled back, when executors are idle, and after a period of idle time (AutoSuspend Timeout), the Virtual Warehouse is suspended.
You set an AutoSuspend Timeout to configure how long a Virtual Warehouse idles before shutting down.
AutoSuspend Timeout is independent of the auto-scaling process and only applies to the original Virtual Warehouse and not to any additional warehouses that are created as a result of auto-scaling.
If a query is executed on a suspended Virtual Warehouse, then the query coordinator queues the query. When a queued query is detected, an executor group is immediately added to run the query on the Virtual Warehouse.
This feature can help lower cloud costs.
Set AutoSuspend Timeout to the number of seconds you want the Virtual Warehouse
to idle before shutting down.
For example, if you set the timeout to 300 seconds, the Virtual Warehouse suspends itself after 300 seconds to save compute resource expense. The suspended Virtual Warehouse restarts as soon as you run a query.