Data hub - What Data should reside in Data hub


Hi Team,

New to Anaplan Modeling. Based on the training, its my understanding that data hub is important to act as source of truth for data that is shared between multiple spoke model. I see data hub as a repository for list, hierarchies and data modules that are shared in more than one spoke models.

My question is what about data that are unique to spoke model or solve a unique use case? Does that also reside in data hub or should it directly be imported into spoke model?


Best Answer

  • rob_marshall


    That is not what you stated…You stated the data is unique to that model and will not be shared with other models.

    If the data is shared, land it in the data hub and then farm it out to the model(s) that need it.


  • @sagarkpr

    Take a look at this article if you haven't already, I believe it will help your understanding…

  • Thanks this article was too helpful but here is a use case and can you advise what makes more sense

    One of the spoke model needs a data that is unique to that model and would not be used in any other model. Should I still import that data into Data hub and then load it into Spoke model?

    Or should I directly create import process to load this data from external source into Spoke model?

    What is the strategy you recommend and pros/cons for selecting it?

  • @sagarkpr

    Why not have multiple data hubs? The data hub is used to land the data, then you push it to the spoke models. If that is one spoke model, all good. In fact, that is the exact method customers use when they only have one use case. You want to minimize the amount of loading into the spoke model and if you are having to load files, build lists, do any transformations, then you are adding to potential concurrency issues.

  • But then having multiple copies of the same data even when they are not shared. Doesnt that add eat into the size of the workspace. I understand the relevance of data hub for things are been shared. Those I can keep in data hub and make sure all the spoke models use the same version of data.

    But data or list that are unique to spoke model, why load them in data hub (add to size) and then import them to spoke model.