GCP Cloud Spanner CDC pushes to Pubsub via Dataflow

I am trying to build dataflow for change data from cloud spanner to pubsub topic, however after providing necessary information when I click on create job, it directly fails and gives error as follow:
"Failed to start the VM, launcher-202305251132348291864736406387748, used for launching because of status code: INVALID_ARGUMENT, reason: Invalid Error: Message: Invalid value for field 'resource.networkInterfaces[0].network': 'global/networks/default'. The referenced network resource cannot be found. HTTP Code: 400."

It is very much unclear if I need to create any VPC since there's no requirement in GCP doc about needing VPC and for what? if VPC is really such important, why GCP doc doesn't mention about it?

By the way, another question If you know, For Cloud Spanner when I am creating Change Data Stream I didn't create a separate instance/DB to store metadata for change since my changes & table won't have thousands of rows, it will be quiet smaller as only text values in 8 columns will it hold. is it fine to have metadaba & actual DB same?

I have figured the answer that I need to provide only subnetwork value and it will take other information out of it.

Regarding Question-2, I can use the same table. GCP will create another metadata table but still while I configuring dataflow job I should mention database name and metadata database name as same table, not the one GCP created.

It is recommended to use a separate database for the metadata table in the change stream pipeline (https://cloud.google.com/spanner/docs/change-streams/use-dataflow#metadata). If you are using the same database, make sure the change stream is not tracking the metadata tables created for the pipeline, since that will cause unnecessary generation of change records.

In addition, you should not pass in the metadata table name. The Dataflow pipeline will create the metadata table for you automatically.