Skip to main content
Documentation
Feature Requests Trust Center Status Page

Community

Destination Improvement: Redshift SUPER type

Answered
Kevin Chao User

Redshift recently announced general availability for the SUPER type (JSON). Currently, Fivetran converts JSON into a varchar type, but would love to see these converted to the new SUPER type.

Please sign in to leave a comment.

Comments

8 comments

Date Votes
  • Official comment
    Kevin Kim Fivetranner

    Hi Kevin!

    Kevin from the product team here. The SUPER type is definitely something we are looking into. While we don't have a committed timeframe, I will discuss this with our Redshift engineering team to see when we can support this new data type that Redshift offers.

    I agree that SUPER is the correct choice for JSON data, and we are happy Redshift has announced this new type! It will only be a matter of time until Fivetran supports this. Please stay tuned for more details.

  • Felipe Rodrigues User

    Hy, have you some update on that? Can I change the colum type from varchar to super on columns with json data?

  • Ayal Alon User

    Hi team,

    it's been 2 years since Kevin Kim response.

    any updates?

  • Darek Deszczynski User

    I would like to know as well if there is any progress on this.

  • Gary Tang User

    +1 on this. We have a dynamodb connector and all the key-value pairs are stored as a varchar. It would be much easier for us to store these with Redsdhift's super data type

  • Edward Miller User

    Indeed, and this would also be a preferred conversion for text BLOBs from other databases, such as the text or ntext types in MySQL.

  • Darek Deszczynski User

    This feature is currently supported for select accounts. Ask your sales rep to enable it in yours. There is a major issue with it though: replication breaks once any of the values inside JSON exceed the max varchar size. Please upvote the feature I created: https://support.fivetran.com/hc/en-us/community/posts/15792576434455-Destination-Improvement-Redshift-destination-Super-column-needs-the-TRUNCATECOLUMNS-option

  • Chris Gilbert User

    This feature isn't currently available. Unfortunately we have hit the varchar size limit on one of our tables, which isn't particularly big at 64KB or so. Super columns allow up to 16MB (though individual strings must be below the 64KB limit still). It would be really helpful if fivetran could support super datatype, as it is the correct way to use JSON in Redshift. As it stands, we are either going to have to add another sync process for our table, or switch to a different product.