Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
Hello
We have a set of QVD files with the same structure - one file per month. Qlik Sense crashes reproducible loading these files.
This is the script:
pos:
LOAD * FROM [lib://Data/pos_2013-*.qvd] (qvd);
This is the output:
pos << pos_2013-01
(QVD columnar optimized)
Lines fetched: 132.498
pos << pos_2013-02
(QVD columnar optimized)
Lines fetched: 262.638
pos << pos_2013-03
(QVD columnar optimized)
Lines fetched: 413.367
pos << pos_2013-04
(QVD columnar optimized)
Lines fetched: 570.108
pos << pos_2013-05
(QVD columnar optimized)
Lines fetched: 732.856
pos << pos_2013-06
(QVD columnar optimized)
Lines fetched: 909.663
pos << pos_2013-07
(QVD columnar optimized)
Loading QVC file: 138.909
After the last line Sense Desktop crashes to desktop. With Sense Server 1.00 the engine crashes.
The same happens when loading the files individually and explicitly or implicitly concatenating them:
pos:
LOAD * FROM [lib://Data/pos_2013-01.qvd] (qvd);
Concatenate LOAD * FROM [lib://Data/pos_2013-02.qvd] (qvd);
Concatenate LOAD * FROM [lib://Data/pos_2013-03.qvd] (qvd);
Concatenate LOAD * FROM [lib://Data/pos_2013-04.qvd] (qvd);
Concatenate LOAD * FROM [lib://Data/pos_2013-05.qvd] (qvd);
Concatenate LOAD * FROM [lib://Data/pos_2013-06.qvd] (qvd);
Concatenate LOAD * FROM [lib://Data/pos_2013-07.qvd] (qvd);
Is this a known bug?
Olaf
I would suggest to open a Case with Support to get this checked in detail.
It seems to be fixed in 1.1. But now it is slower than the workaround solution consisting of simply forcing an unoptimized load. Load time went from 30 minutes up to 45 minutes. Unfortunately I have no time researching this deeper in the moment.
Another thing I noticed, using "where exists(fieldname)" resulted in an unoptimized load. I'm sure in 1.0.0 to 1.0.2 this was optimized.
Olaf