Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
Dear Experts -
Currently the Qlikviews application hold relational data (qvw+qvd). These qvd hold some historic data which were processed using relational source.
Now, will there be any issue with existing qvd, as i would like to append the existing qvd to hold data coming from hadoop or do in need to reload the entire historic data into qvd replace with relational.
Kindly advice the best practices or approach moving from relational to hadoop source.
Many thanks in advance -
Sean.
First of all a correction. QVD and QVW files do not contain relational data, they merely contain data, so the form of the data source (relational or semi-structured) has no impact on the final data in the qvd/qvw files. Technically, a Qlik data model is an associative model, not a relational database.
The load script will need to be changed to read the data from the hadoop source and you can apply whatever transformation is required in the load script to produce records structured to match the history.
Thank you Jon for your quick response. Another quick question.. What is the best approach on pull and push mechanism in QV. Are there any additional plug-in required to gracefully document refresh.
Appreciate, if there are any reference links available or any best proven practices available.
Thank you!
SEAN.
Qlikview operates on a pull mechanism. When a reload is triggered, the load script executes, fetching data from the data sources and transforming the data as required. There may be extensions which allow for Qlik to poll a source and fetch new data when available (a sort of push), but out of the box you could only run the pull over shorted intervals. If you search here and/or in google, you should find Qlik white papers and best practice guidelines.