Post reply

The Data Fabric – Future of Data Management

  • February 1, 2023 at 12:00 am

    #4141766

    Comments posted to this topic are about the item The Data Fabric – Future of Data Management

  • February 1, 2023 at 1:33 pm

    #4144186

    Ajay,

    Thank you for posting this. There is certainly a lot of buzz reviving around data fabric. At it's core is a concept that used to be called data federation. Or, more broadly, data virtualization. In the past, I would typically see implementation efforts get bogged down by  a few key aspects.

    1. Source systems struggling with the weight of real-time analytical queries.
    2. Virtualization tools producing inefficient queries that frustrate users.
    3. Schema and query path maintenance to keep up with evolving data structures in the source systems.

    Have you been able to successfully tackle any or all of these obstacles? With the additional components involved in data fabric, can these be accounted for in an automated way? Also, is there a balance of copying some data while virtualizing other data that doesn't result in doubling the investment in the overall system?

    Just some questions I have that I'd love to get real-world perspective on because I'm excited to see this field progress further along.

    Thanks,

    Phil

  • February 1, 2023 at 3:20 pm

    #4144259

    Hi Ajay,

    Thank you for sharing this article!

    I have three questions:

    • How are the "Golden records" identified by Reference & Master Data Management solutions positioned and how are these connected in the Data Fabric?
    • How are transaction records regarding various entity types, originating from various data sources and having a different grain connected in the Data Fabric?
    • How is data in historic perspective connected (when the organization has not only information needs matching the current point of view) in the Data Fabric?

    Thanks,

    René

  • February 2, 2023 at 5:46 pm

    #4144808

    I should like to see this in operation somewhere as I've heard a lot of talk and seen some very expensive sales pitches.

    It seems like a more ambitious form of data federation, virtual data warehouse etc.

    I think there needs to be a cultural revolution for Data Mesh/Fabric to be a success.  I have dealt with teams that view data as a burden and a curse rather than gold or oil.  I don't think such teams are attitudinally suited to providing a data service.  Until that changes I feel that Data Mesh is an ambitious dream rather than a reality.  I'd love to be wrong.

    LinkedIn Profile

Viewing 4 posts - 1 through 3 (of 3 total)

You must be logged in to reply to this topic. Login to reply