Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
Hello I got a QVW document that has a fact table with 22 columns and 121,000,000 records
The size of the document on the disk is 2.7GB; and when I open the document it consumes about 8.5 GB of RAM
in this document I got only one pivot table with the following expression;
whenever I open the pivot table (initially collapsed), the document inflates and consumes aobut 19GB of RAM
is this normal? Actually I'm unable to continue with the development as the document stalls several times
even if i filter the document the calculation takes a lot of time; I need to wait about 1 minute for the expression to return results
the expression is as follows and is used to calculate the effective retail:
sum
(
Aggr
(
FirstSortedValue
(
{
<
[Transaction Date]={"<=$(=vSelectedDate)"}
>
}
[Amount USD w/o VAT],-[Transaction Date]
)
*
sum
(
TOTAL
<[Geo Country],Brand,[BU Code],RetailKey>
{
<
[Transaction Date]={"<=$(=vSelectedDate)"}
>
} Quantity
)
,[Geo Country],Brand,[BU Code], RetailKey
)
)
Hi Ali , There is , in between this statement pls check [Amount USD w/o VAT],-[Transaction Date].
Regards,
Alvin.
Sry there are also syntax errors in the expression where you are using the set analysis . Please check .
Regards,
Alvin.
I fixed the error in the post above
however this is not the problem the expression is giving correct results
yet the response is extremely slow
when the chart is collapsed the document only consumes 8 GB of RAM
when I expand the chart the document consumes almost 20GB why?
can we optimise the expression?
thank you
Hi Ali, Pls try to left join the dimensions with the main table .
There will be increase in the Reload time .. But hope fully the Ram Consumption will be less. And also try to group it at the Script level.
Lets give a try ....
Regards,
Alvin.
Hi
Do all the fields in the expression come from the main table? Or are they closely associated (one jump to the main table only). If not, then QV is having to do a lot of work to perform the filtering, leading to poor performance and memory bloat.
Bringing all the fields into the main table (or no more than 1 jump away) may improve performance.
Regards
Jonathan
ok I'll try to make all the columns be in the fact table using joins
but is there any difference whether these columns are key columns or not?
Please advise