SSAS Tabular delta schema changes and processing
Hello Experts,
We are using AWS hosted SQL and SSAS Tabular both on same instance and the plan is to build environments as
Dev & Test Environment: 3 & 6 months respectively,
UAT: 1 Year
Pre Production: 5 Years
Production: 10 Years
Our product is still maturing which as a result frequent schema changes (either existing DAX measures, calculations, relationships are being modified, or new measures/relations are being introduced) are occurring, and the updated build DAX code is being deployed to Dev/Test environments. Currently whenever such deployment request came, after connecting to SSAS from EC2 machine, first I drop the existing dev/test tabular database, and then follow the AWS process to deploy the tabular database(copying files .asdatabase and .deploymentoptions to S3 bucket, download and deploy) by calling the below stored procedures
exec msdb.dbo.rds_download_from_s3
exec msdb.dbo.rds_msbi_task
@task_type='SSAS_DEPLOY_PROJECT',
this creates a new Dev/Test blank database where further reprocess the data after making connection to respective SQL datawarehouse and load data. Since deployment is frequent and deleting then re populating data into tabular is time consuming, therefore ...
Is there any way which should help to keep the existing historical data as is and only delta changes are deployed to Tabular (delta schema in visual studio)? and only impacted data is being refreshed?
The question is coming from the perspective of building higher environments where the challenge is to keep the historical data and only process Delta schema changes and respective data in tabular.
Could you guys please elaborate and suggest the industry process for such data projects?
Thanks in advance!
Prashant Pandey