DR for OSS prefect servers

[zzstoatzz]

converted from #17318 by @HeavenS221

Describe the current behavior

no disaster recovery method on prefect server

Describe the proposed behavior

We hope prefect server have some disaster recovery method when using multi kubernetes clusters.

Example Use

For example, we decided to use prefect as task orchestration tools. There is a requirements for remote disaster recovery.
Assume we have several clusters in different regions, each with a Prefect worker, and one of the clusters hosts the Prefect server. If the cluster with the Prefect server fails, we want a method to quickly deploy the Prefect server to the other clusters, therefore the existing worker can connect to the new prefect server. So that the Prefect services can be restored as quickly as possible.

Additional context

I didn't find any info for these things, any use suggestions or best practice would be great.

Best regards

[HeavenS221]

thanks for convert it into discussion, we can discuss it here.

[HeavenS221]

@zzstoatzz may I ask if prefect-server is stateless？