Disconnected buckets can be re-connected, but then servers need to
be re-authorised to access the bucket.
Warning: The disconnect will only be applied to the live instance
on the next deployment. To more quickly disconnect, disable the
service account that was used to authorise access on Google Cloud's
UI.
Bucket ID:
Tip: You can use the same service account to for multiple buckets or
use multiple service accounts for the same bucket (depending on your
security requirements).
Make up an internal name memorable that is unique and easy to
remember, this is what you will use to specify which bucket to
access with a pcf.
Internal Name:
Bucket Name:
Authorisation JSON:
Note: The testing bucket name is "testing_" + your live bucket name, e.g. if your live bucket is called "My_Bucket", then you must also have a testing bucket with the name "testing_My_Bucket".
Authorisation JSON (testing):
Object storage is lower-cost bulk storage that is independent from your Kubernetes (K8s) deployment. This is the same type of storage used for your PostgreSQL relational database backups and for copies of your codebase from past deployments (make sure to use a different bucket though, the bucket used for system data should not be added here).
Storing data separately from your K8s deployment increases data safety, because if you K8s cluster is lost, the data should not be affected and thus still be accessible.
Tip: Carefully review the setup details of your bucket. You can set up data policies (including bucket policies, minimum retention policies, and even the region to store your data in) to increase data safety to the degree necessary for your organisation.
about