What is faster? Resident Loads or Loads on Preceding Loads?

[mwallman]

Hi,

Can anybody please tell me what is faster? Resident Loads or Loads on Preceding Loads?

Or generally they are about the same?

E.g. I want to do some further calculations on some fields, is it better to use Resident Load or Preceding Loads?

[Peter_Cammaert]

Yez.

[swuehl]

Sunny, just read the second and third comment from

Preceding Load

[#4]

I ran test script on 10 milion rows (random test data).

Optimized load 6sec.

Preceding load (with 1 'if' and 'multiplication') 23sec.

Resident load with the same transformations as above 11sec.

But from my short experience it looks like 'it depends' on where conditions and transformations used.

[Colin-Albert]

My two penn'orth....

Preceeding loads are faster than resident loads (usually), but in some cases you cannot use a preceeding load.

If you use a wildcard in the from; or if you have a crosstable function, then you cannot use a preceeding load

Sometimes I have found that storing a table to QV, dropping the table and then and reloading from the QVD can be faster than using resident load, even though the QVD load is not optimised.

I don't think there is a simple "one rule always applies" fix, simply if you think the option you have chosen is not performing well, then try another approach!

[Clever_Anjos]

Comments from rwunderlich‌

The Cost of Preceding Load | Qlikview Cookbook

[Lech_Miszkiewicz]

I think Rob just reviewed this few months ago, since there were changes in performance over last few years:

This shows it all

Preceding Load Performance Update | Qlikview Cookbook

regards

Lech

[Clever_Anjos]

yes, according to him there´s no noticeable difference into latests versions