In order to help users diagnose issues around serverless lock errors, we should log the following when failing to acquire the lock :

• The type we're trying to acquire the lock for
• The operationId we're trying to acquire the lock for
• The type that already hold the lock
• The list of operationIds that are already active

We should also improve the logging of TenantMigrationDonor to say "failed" rather than "Tenant Migration completed".