We are migrating our DW from postgres to snowflake. Our quicksight instance contains both custom SQL and direct tables. Any tips or resources on the best way to approach this? All scripting approaches have resulted in so many errors in the analysis and dashboard files (primarily due to calculated field errors). We have 100’s to migrate and are trying to avoid manually switching datasets in each one.
Hi @carysylvester - One of the approach is to replace the data sets in the analysis via scripting. If the field names and data types are same, then you do not need worry about the calculated field. However it is better to do a POC before cascading in all analysis.
Regards - Sanjeeb
Does Sanjeebs approach seem like something you could attempt?
Like he said, if schema the same, new datasets could be created + analysis updated programmatically.
1- Create new datasets using original datasets (update JSON from describe dataset to use new custom SQL and new Datasource). Schemas must match or otherwise schema has to be update.
2- update all analysis and dashboards with their respective new datasets.
Happy to explore the scenario further or share some partners - via DM - that could help your with the migration if that is an option.
Thanks,
Ramon
1 Like