Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
Hi,
I recently developped an app and some field names are quite long. Q1: Do you think it can have an impact on the app performances?
Due to the app requirements (app in an Iframe, Website look'n feel), I used extensively the Hide/Show functionnality for charts and tables. There is nearly 400 objects in some sheets but only about 10 are showing at a time depending on the selection. Qlikview tabs are not visible in the Iframe and I use buttons/listbox to mimic the Qlikview tabs.
It takes about 20 seconds for the app to load on the AJAX Web client but because the app is in an Iframe, you cannot see the "loading" progress.
The app reactivity is not perfect: takes about 2-3 seconds to show/hide charts when you select a tab or a type of data in a listbox.
The server we use is a Quad core hosted VM with 8Gigs of RAM and the app weights about 12MB only...
Q2: What would you suggest for the app to be more performant/reactive?
EDIT: I found a version of Qlikview Optimizer and QlikviewServerPerfrormance that I attached to this post.
Hi,
if only 1.5 GB is being used, memory is not your problem (at the moment, even though you are getting pretty close). Try looking into the following:
- Open Document Properties, Sheets, and look at the sheet objects list in the bottom. Notice the objects that take most time / memory to recalculate and try to analyze what's especially heavy with those objects.
- Find "QlikView Optimizer"- it should be available in "Share QlikViews", and load your memory statistics there. Analyze your data and screen objects for the heaviest performance hits.
- One problem that I learned the hard way is that, when dealing with large data sets, QlikView seems to perform better with fewer objects on the screen (even if each object has a higher number of dimensions and expressions) than with lots of small objects (like for example numerous text objects and small sparklines on the famous Financial Controlling Dashboard). If you can convert a number of smaller objects into a single bigger object - you might gain in performance.
- If you have any macros, try to avoid those at all cost. They are hitting performance very hard on a large data set
- Watch for heavy expressions, especially involving IF statements. No matter what people might say on the forum, IF statements are prohibitive for large data sets.
good luck!. otherwise refer this link
Hi.
1.Howmuch possible avoid if condition in Expression.
2.if any calculation need first try to calculate in script level,not possible do in frond end. use simple expression in graph.At time show one graph ,remain all graph minimize.when you more graph graph maximized this time application performance will slow.
3. unwanted fields comment in script level.unwanted field also occupied memory.
Regards
Perumal A
Hi,
My dimensions and expressions are quite simple appart from some charts where I use conditional expressions. nothing too fancy.
Most of my calculations are done at a script level and all the fields loaded are in use.
I guess I should try to split the sheets into different tabs somehow because 390 objects in a sheet is too much maybe, although they're hidden.
I also tried to make a "Landing page" with 1 object on it to see if it would boost the loading time but it did not...
Any other suggestions?
Hi.
1.your thoughts is correct.first split sheets based on your requirements.
2.when you are applying show/hide condition in graph,same condition put in Calculationcondition place in graph general Tab.otherwise your graph always refreshing in sheets.
3.How much possible avoid text objects.
Yes that's what I thought... I already have calculation conditions on the graphs + show/hide condition.
Why should I avoid text objects?
Hi,
if only 1.5 GB is being used, memory is not your problem (at the moment, even though you are getting pretty close). Try looking into the following:
- Open Document Properties, Sheets, and look at the sheet objects list in the bottom. Notice the objects that take most time / memory to recalculate and try to analyze what's especially heavy with those objects.
- Find "QlikView Optimizer"- it should be available in "Share QlikViews", and load your memory statistics there. Analyze your data and screen objects for the heaviest performance hits.
- One problem that I learned the hard way is that, when dealing with large data sets, QlikView seems to perform better with fewer objects on the screen (even if each object has a higher number of dimensions and expressions) than with lots of small objects (like for example numerous text objects and small sparklines on the famous Financial Controlling Dashboard). If you can convert a number of smaller objects into a single bigger object - you might gain in performance.
- If you have any macros, try to avoid those at all cost. They are hitting performance very hard on a large data set
- Watch for heavy expressions, especially involving IF statements. No matter what people might say on the forum, IF statements are prohibitive for large data sets.
good luck!. otherwise refer this link
Thanks. All of this makes sense and I shall apply this in the future.