App size is > 5 gb

[Tarun_Qlik]

when I'm fetching data from one table. app size is 2 GB and after doing transformation and joins with other two table it is > 5 GB. I have tried to optimise it with autonumber , timestamp and by removing unnecessary fields but no luck.

(App size--> 2 GB in below scenario)

ABC:

Load *

;

Select * from table;

Table has 3 years of data (13 Million records)

[PrashantSangle]

check, if you can use Dynamic view. 

So that on your dashboard, you can use only summarized view data and if required fetch detail data from source for specific selection.

Regards,

Prashant Sangle

[Tarun_Qlik]

actually there are few reports where I need to present Detailed data.

[marcus_sommer]

Maybe the joins are adding extra records and the final table has a lot more data. To get more specific hints for possible optimizations you will need to share more details - to the data-model, the kind of data and transformation ...

[Tarun_Qlik]

ABC:

Load *,

Order date,

shipped date,

return date

;

Select * from table;

all 3 dates are coming from a single table. I have merged them into one and added a flag for each date (Concatenate using resident load). Final count for that table after concatenation is close to 39 million records.

with other two table it's a simple left join.

[Tarun_Qlik]

My First ask is ' when I load only one table with 13 milliion records before making any changes(except autonumber , timestamp and removing unnecessary fields)' app size is 2 GB.

[marcus_sommer]

2 GB for 13 M records is already quite big - whereby with rather wide tables and several fields with many distinct field-values it might be the smallest possible size. Nevertheless I suggest to look if any of them could be adjusted - especially splitting the timestamp, removing record-id's, removing row-level formatting and similar things.

By concatenating these data 3 times - for each date-field - should not much happens in regard to the file-size because just the row-id of the data-table increased. Therefore something else happens - maybe the concatenating enforced now a row-level formatting whereby I doubt that this would explain the grow from 2 GB to 5 GB.