For my Cloud Run service, Both RAM and CPU are low. I don’t know why GCP has allocated 6 instances. Its increasing my bill.

CPU is nowhere near the 60% required for scaling.

RAM and CPU:

Instance count:

Interestingly, I have been reducing the CPU and memory limit and the instance count remains unchanged. This has reduced my bill.

I changed from 8GB RAM, 4 core -> 4GB RAM, 2 core -> 1GB RAM, 1 core.

I do not want to goto 512 RAM, <1 CPU.

Is this some dark-pattern by GCP, to increase my bill while I can’t do anything about it?

Other config:

General

CPU allocation:
CPU is only allocated during request processing

Startup CPU boost:
Enabled

Concurrency:
80

Request timeout: 300 seconds

Execution environment: Second generation

Auto-scaling

Min instances
2

Max. instances
25

CPU limit
1

Memory limit
1GiB

Its a nextJS server, mainly serving json API calls.