Qlik Community

Ask a Question

QlikView App Dev

Discussion Board for collaboration related to QlikView App Development.

Announcements
Support Cases coming to Qlik Community Oct. 4! Start chats, open cases, explore resources: READ DETAILS
cancel
Showing results for 
Search instead for 
Did you mean: 
prahlad_infy
Creator
Creator

What shall be the possible steps to reduce qvd creation time from another qvd ?

Hello Sirs ,

In the below scenarion What shall be the possible ways to reduce qvd creation time from another qvd (STEP 2)  ?

qvddddddddddd.png

8 Replies
jfkinspari
Partner
Partner

When you don't write the purpose of QVD2 nor the steps performed when it's created, I can only come up with a few generic options
1) Make an QVD optimized load of QVD1
2) Increase hardware specs for the machine performing QVD2
prahlad_infy
Creator
Creator
Author

Hi Jfkinspari ,
QVD2 is getting created from QVD1 after applying few business rule based transformation on QVD1 .

Apart from above two option , what should be other considerations while scripting to reduce time .
Please let me know all possible that comes to your thought .
jfkinspari
Partner
Partner

You could find out which transformations is actually taking the most time --> where is the biggest potential. In case you are making a single LOAD with a few transformations inside, it could indicate the issue is related to slowness in reading and writing the file. Unless it contains billions of rows.
prahlad_infy
Creator
Creator
Author

Sorry , if i am asking a very simple question . 
Is there any way to track and note which piece of script is taking more time , other than manual eye rolling on window .

 

jfkinspari
Partner
Partner

No, not really any other method than doing it manually.
You could script some TRACE statements with the log file, that later could be used as anchors for analysis, bit is probably overkill for your need.
prahlad_infy
Creator
Creator
Author

Thank You Jfkinspari 
Being new to qlik , i often end up asking comman things like this .
Now i need to look for , how to use TRACE statement .


let say there is a qvd load and i am applying two lookup and one subfield statement , would TRACE help me in finding how much of time each lookup and subfield took ?

 Load
f1,

f2

,f3

,lookup(............) as F4
F5,

F6,
lookup(.............) as F7, 
subfield(..........) as F8 

from xyz.qvd ;

jfkinspari
Partner
Partner

TRACE just makes a entry in the log, which you could look for when analyzing the log.
I suggest you simply look for the timestamps in the log file in order for you to find out which LOAD statements is consuming the 1 hour.
marcus_sommer
MVP & Luminary
MVP & Luminary

Most effectual would be to use incremental load-approaches on each of your transformating-steps. Further if your extractings contain multiple load-statements it might be useful to separate them into multiple working-steps, too. In the second half of the following posting you will find various examples to incremental approaches and exists() which is needed to keep qvd-loadings optimized:

Advanced-topics-for-creating-a-qlik-datamodel

- Marcus