Repurposing this epic to be more tightly scoped to our "discovery service needs". Through our architectural design process, we've become increasingly more confident that a CRD surface + controller as an abstraction layer between the MTX UI and the pipelines is not necessary.

The operator should be able to deliver a collection of ClusterTasks that wrap the various steps required to sequence a migration. The UI can build up a pipeline composed of these tasks dynamically as a function of the input provided in the wizard by the user.

However, there still appears to be a need for some type of backend service that is capable of providing several functions, in service of a richer UI experience:

• We want to be able to present to a user a list of namespaces that they have access to so they can choose from that collection, and not require them to type in the name of the namespace they wish to migrate
• We want to be able to provide "MigAnalytic" type information, such as number of objects in a namespace, real PV usage statistics (how much capacity has been used in a volume), and PV discovery (this is the list of PVs in the namespace you have specified)
• We want to be able to perform "deep validations". This is something that goes further than a shallow validation of, say, 63 char limit, and requires us to do things like confirm credentials provided are valid and work, or to confirm that a file can be uploaded to an S3 bucket given the credentials provided.