This site uses different types of cookies, including analytics and functional cookies (its own and from other sites). To change your cookie settings or find out more, click here. If you continue browsing our website, you accept these cookies.
I don't know Qlik Views ETL specifically but I came to Alteryx from both programming and SQL Server Integration Services.
I used it in anger for two weeks and it changed how I thought about this kind of thing. At first, I thought ah good prototype tool that would let me work out ETL and process and then I could code it up.
As I used it, I changed the way I thought about using it. I was able to involve buisness owner and respond within minutes to changes. Huge and very critical processes ended up being both easy to understand and very quick to run.
Nowadays, I find myself wanting to extend Alteryx and do more and more in it. I love picking up a challenge from the community and seeing what is possible.
May I answer my own question? After all, experience is the best teacher.
Qlik has its own ETL engine, but it is pretty much limited. Using Qlik we need to perform standard SQL queries to alter and transform the data, and it might not be so easy if the source data is full of garbage.
Consider we can bypass the hardships in querying using Qlik, we still need to face another factor and that is: compile resources. Each script coded in Qlik takes resource and time to load. The more complex the ETL process, the longer it will be. Additionally if we have a very large datasets, it might disturb the in-memory techonology QlikView employs.
QlikView was fated to help in data visualization and it works the best in it, do not burden QlikView with too much scripting and ETL process. Give QlikView what it wants, i.e. visualization. So, where do the ETL processes go to? Alteryx will be the prince riding a white horse.
Alteryx transforms all data sources into medium-rare or even well-done data steaks for QlikView to eat. With QVX output from Alteryx, QlikView can easily read it and visualize it. We cut all the ETL processes and resources in QlikView, therefore allowing QlikView to perform even faster. No more page load.
Alteryx do all the dirty work, even we can cleanse the data easily using Alteryx.
I have not yet compared which is faster in the time performing ETL between the two, but I believe Alteryx is faster if we can put the right process flow, and it is also easier, preventing us from scripting errors.
Usually we just dump all the data to qlikview with time = t1.
But since we use Alteryx and the reload it to qlikview, it will take time = t2 in Alteryx.
Therefore to use Alteryx and Qlikview in an environment with minimum utilization, we will take more time... (t1 + t2).
However is this bad?
Have anyone experienced a scenario where Qlikview takes so long to switch between one view to another view? Long loading time during filtering? Not responding? This scenario occurs when we use Qlikview to do all the ETL. Therefore even when loading the data takes less time than using Alteryx, our actual data-digging will take longer. Getting the answer to your question takes forever.
Therefore if we let Alteryx to do ETL first and dump the data to Qlikview, we can avoid the slow qlikview response.
Afterall, performing ETL with Alteryx does not take that long...
if as above t1 = time to load data to Qlikview directly.
t2 = time to load and perform ETL in ALteryx.
Then, I can say that t1 > t2, therefore using Alteryx before Qlikview will not result in 2*t1.
Let the flow take a bit longer, but it will be worth the while...
and last but not the least, it does not always take longer.