Model change has been rejected to prevent the workspace exceeding its size limit

kiwi.chow@huhtamaki.com
kiwi.chow@huhtamaki.com
in Integrations and extensibility

Hi,

 

I am working on a Sales Forecast Model by using customer and item for a Region.


In the Data Hub,

number of customers = 13,083

number of items = 24,876

 

However, I got an error message when I tired to create the forecast model by using customer and item (see attached.)

 

Is there any memory alert in Anaplan or how can I fix this issue if I want to create the model in this level??

 

Thanks,

Kiwi

Tagged:

Best Answer

  • Kanishq17
    Kanishq17
    Answer ✓

    hi kiwi.chow@huhtamaki.com 

     

    The module you are trying to create will cause the size of model to be so much that the workspace in which model is there will go over its specified limit, so workspace has a limit and sum of size of models making the workspace must not go over the limit, i.e it is preventing to create the module, also creating such a huge module is not feasible for performance issues as sparsity maybe there.

     

    What you can do is create a numbered  list by concatenating customer and items which are to be sold under a region with region as its parent, and use this list as a dimension in the module

Answers

  • nathan_rudman
    nathan_rudman

    the module you are trying to create is bigger than your model by 6 order of magnitudes.

    Space used is a simply calculation of the number of item: 8 bits (for number format) x size of list 1 x size of list 2 x ... x size of time x size of version (if applicable).


    Surely you don't need all those dimensions. Reduce the number of dimensions to have ONLY what you need, by line item.

    Split into several modules as some line items will not need the same dimensions as other.

  • JoeAnderson
    JoeAnderson

    Agree with Kanishq17.

     

    We've used this strategy numerous times to see massive size savings. If it's like a lot of data sets, the dimensions "customers" and "items" are sparsely populated with valid data.

    We've concatenated as many as 4 items to create a unique key (maybe more...)

  • einas.ibrahim
    einas.ibrahim

    Hi kiwi.chow@huhtamaki.com 

    All the responses you got are excellent. Yes, your available workspace is smaller than the model-to-be if you apply these dimensions. 

    My advice is, even if you had the space required, to not proceed with the lists as is. Unless each customer will be associated with all or the majority of the items then we shouldn't create all these intersections by dimensionalizing by both lists.

    The idea of creating a combined list is a good idea, but also, look at lists and ensure that all items in the lists are valid (no inactive customer for example or duplicate items,..etc). Also, you can look into grouping the items and/or customers into hierarchies and plan/report on that level if your business case allows for that. then drill deeper when needed.

    Good Luck

  • kiwi.chow@huhtamaki.com
    kiwi.chow@huhtamaki.com

    Hi Kanishq17,

     

    This idea is very useful and my mind is quite clear how to proceed now.

     

    Thanks you so much !!

     

    Hi All,

     

    Thank for your replying, your input is also quite helpful : )

     

    BR,

    Kiwi

Categories