Update schema of existing Dataset

[spranati]

Is your feature request related to a problem? Please describe. I need to update the schema of an already created dataset using the config.json. I tried updating the schema, changed the schema name so a new schema is created and associated, however, when I drop the updated config.json in S3, I get an error that the dataset is still referencing the previous schema and not the one I provided. Seems like it is creating a new schema but not associating with the dataset. Although I am able to update the schema directly from console/CLI, this functionality is not supported by using config.json. It would be nice to be able control this directly from config.json instead of using another way to manage on of the components. Describe the feature you'd like Update schema of existing Dataset using config.json. When I update the schema of in the config.json and provide a new name, it should be able to create the new schema and associate it with the existing dataset that I am referencing.

Additional context Error:

{ "errorMessage": "expected schemaArn to be arn:aws:personalize:us-east-1:288364316505:schema/bell_dev_vod_users_schema_v2 but got arn:aws:personalize:us-east-1:288364316505:schema/bell_dev_vod_users_schema. This can happen if a user modifies a resource out-of-band with the solution, if you have attempted to use a resource of the same name and a different configuration across dataset groups, or are attempting multiple solution maintenance jobs at the same time", "errorType": "ResourceFailed", "requestId": "a80171af-31de-4b1a-80b5-9acfeac78ce2", "stackTrace": [ " File "/opt/python/aws_lambda_powertools/metrics/base.py", line 411, in decorate\n response = lambda_handler(event, context)\n", " File "/opt/python/aws_lambda_powertools/tracing/tracer.py", line 305, in decorate\n response = lambda_handler(event, context, **kwargs)\n", " File "/var/task/shared/sfn_middleware.py", line 350, in decorator\n self.check_status(resource, **kwargs)\n", " File "/var/task/shared/sfn_middleware.py", line 304, in check_status\n raise ResourceFailed(\n" ] }

[fhoueto-amz]

Thanks @spranati. We will review and come back

[fhoueto-amz]

Hi @spranati Based on our discussion, our PMT will continue the review and we will update if we put it on our roadmap or not

[spranati]

@fhoueto-amz any update on this ?