We took a benchmark on GCP S3, but the results were strange for us. The number of IOPS for the request put of a 4K object reaches a maximum of 100.
We ran benchmark client on GCP compute vm and apparently it was not bottlenecked anywhere.
vm configs:
- Region: us-centeral1 (lowa)
- zone: us-centeral1-a
- Series: E2 Machine
- Type: e2-standard-8 (8vcpu, 32 GB memory)
- Disk Size: 250GB
This is despite the fact that we were able to get much better results from Digital Ocean.
Is there a specific reason for this? any one has idea?