0 votes

I am discovering DSS and this is my test use case:

 

1- Connect to a single PostgresSQL table ( around 4 millions record)

2- Add a recipe to add a column (XYZ)  and output the managed filesystem.

3-Create a pivot table to calculate an average of a mesure ( speed)  ( 1 dimension( the new XYZ) - 5 possible values)

4- Get the answer based on the full dataset

 

So I build the recipe and then go to the dataset=>chart OR the analyse=> chart to design the pivot table.  It works fine with the 10000 row sample but, when i ask for the full dataset sample ( DSS engine) .  It throws an error 500 after  5 to 15 minutes.

 

Any idea?

 

asked by

1 Answer

0 votes
Hi,

It would help to have the "run/backend.log" file. Probably the helper process which manages the chart experienced an out of memory issue, which could happen if some of the column have very large number of unique values.

For doing charts on such large data, we'd recommend that you switch the charts to "In-database" engine (in the Sample settings). This will push almost all the computation down to the database and will provide better performance
answered by
The dataset is the local file system managed (CSV) so no database engine available.  It says in the doc that The DSS Charts Engine does not require that the chart data be loaded in memory, but is instead able to efficiently stream data from disk and perform queries on the fly. This allows you to perform visual analytics on very large data extracts that would not fit in RAM using commodity hardware.  

How can I use this  Charts Engine for my chart ( or pivot table)?
991 questions
1,024 answers
1,075 comments
3,140 users

©Dataiku 2012-2018 - Privacy Policy