Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
Hey People,
New to Qlikview.i have a data of 10 gb for which i have to build a tool on Qlikview. Can Qlikview handle this much amount of data?
How much would the speed of the tool be affected?
Hi,
it depends on many things:
- hardware (RAM, cores)
- data model (one table, many tables, associations)
- cardinality and sparsity of the data (esp. for compression ratio)
- record count
- amount of UI elements (listboxes, charts etc.)
- amount of users
The amount of raw data gives not a real indication. But, I think 10 GB of raw data is not a big size.
- Ralf
We've been using tens of GB of reasonably complex data without too much trouble. Main issue is having enough RAM in the web server, as each additional user of the large QV application will need a fair amount of RAM for her/himself. I recall ca 10% of application size per user being mentioned somewhere?, I.e. for an application using 30 GB of RAM, each user using it will use another 3 GB of RAM (assuming that 10% number is correct).
But there is a lot of performance to gain by designing the data model and application properly, doing set analysis over lots of rows in a chart is a no-no from performance point of view. Much preferred (if possible given the problem at hand) to do the aggregation (or whatever needs to be done) in the load script.
BTW, we're processing 250 GB raw data every day in one project..
But you're not juggling around 250 GB in RAM, I assume... ?
Agreed, QV is pretty effective at quickly ingesting large data volumes, at least if you avoid reading via ODBC (which in my experience is way too slow for larger data volumes).
CSV files and similar are on the other hand quick to read.
/Göran
Well, goran thanks for reply and i had started importing 1.8 gb data and i couldn't..
Data is being fetched but the reloading process continues for a large amount of time without giving any result.
And yes i have a 4 GB ram!!
Help!!
Unfortunately we have only 144 GB RAM but I've heard from a project using 1 TB RAM !
Vanaika,
are you running QlikView on 32bit? If so, it could only use 2 GB per application. I presume it would work on 64 bit.
Are you loading just one table? I think we would need more information to help you.
- Ralf
Vanaika: You could also try loading a subset of the data, to make sure the load script itself works as expected.
You could do this either by creating a small(er) input file to read from, or by using the built-in debugger's limit feature (allows you to load max n lines in each load statement).
Inserting trace statements in the load script also helps pinpoint where the script hangs or has issues.
Ralf: 144GB isn't too shabby though. Given the development of RAM prices you'll soon be crossing into the next order of magnitude... \o/
/Göran
Hey Goran,
Could you please tell me how and where to insert trace statements in the load script?