Large transaction data imports with Code > 60 characters

david.savarin

Hi all,

I'd need some experience sharing regarding some challenging data imports that have the following : 

- Large volume of transactions

- Multiple attributes (> 10) to load

- Unique code would be > 60 characters if we want to follow planuals and best practices to have a good performance load

 

Has anyone solved this challenge ? If that so, how do you perform good performance x large transactions volumes x multiples attributs that dont "fit" into the 60 characters limit of Anaplan ?

lkmishra04

For code >60 you need to organize your data combination to make unique key / code less than 60.

david.savarin

Unfortunately this is not always an option. Acting on system generated ids isnt something you can easily act upon given how structural unique ids policies are within transactional it systems.

kevin.cho

Can you hash your current combination key (>60) char to produce a code that is <60 char (e.g MD5 hash), and then load the combination key as another single line item? That way, you can still break down the combination key into its components, as per DH best practices.

david.savarin

Yes, that s what we ended up doing, given the code limitations as opposed to other text line items.

- create id code, then hash it with md5 which generates a 32 bit hash code and use it as the code for anaplan list ids purposes

- we keep another line item with all the original codes attributes to load these into another line item and calculate the attributes from there.

I am curious if anyone had other solutions