DB extract & reload

Gary Beardsley

Hi Board Platform, do you know how can I clean up db sparsity, if I have a big application? I have cube extracts going beyond 30 Gb and Board can't load them back in?

Samson Sunny

Hi Gary,

Check for any cubes that has exceeded the 64 bit sparsity limit and are operating in the 128 bit sparsity performance mode due to expansion of the entity max item numbers. There is a trade off in the performance of the application when the cubes go beyond the 64 bit sparsity limit. if there are any cubes in the 128 bit mode, consider limiting the max item number size of the sparse entities in the cube to optimize the sparsity sizing.

Kindly check out the below article on sparsity that explains this well.

https://community.board.com/discussion/16440/sparsity

Also check out the following discussion on the 128 bit vs 64 bit sparsity performance trade off

https://community.board.com/discussion/6620/performance-trade-offs-128-bits-sparse-versus-dense-and-64bits-sparse

Thanks,
Samson.

Kevin Bettenbrock

Hi Gary,

Another thing to look at here would be the amount of data being held in Board. Would it be possible to remove any data that is no longer necessary for the purposes of what the application is doing?

For example, if the use-case is only using data for the past two years, it would be helpful to remove data from previous years to reduce the size of the cubes and data model.

Hope this helps!
Kevin

Fethi ZERARA

Dear @Gary Beardsley ,

if you have some loading troubles even after managing the sparsity then you can create data readers to manage the loading of the heavy cubes Year by Year.

once this is done you can reload the rest of the cube using the reload All cubes action.

Hope this will help !

Bettina Clausen

Hi,

if you want to rebuild a specific sparsity you can also use the "optimize" functionality. That way you can easily get rid of no-longer-valid-combinations. Please note that for the "optimizing" time the db will be set to maintenance mode.

Kind regards,
Bettina