Skip to main content
Announcements
See what Drew Clarke has to say about the Qlik Talend Cloud launch! READ THE BLOG
cancel
Showing results for 
Search instead for 
Did you mean: 
Not applicable

fastest way to load a table

Hi Friends,

I have two table

Table 1      Table 2

ID              ID

Fname       Fname

Lname       Lname

Now I have to create a new table taking all values from Table1 and  values from Table2 where ID not exist in Table1

What is the fastest way to load a table like this

thanks

6 Replies
adamdavi3s
Master
Master

There was a big blog post on loading data on Rob's site.

I've always used QVDs rather than resident load to be honest and I think this shows that is faster as well

http://qlikviewcookbook.com/2013/09/fastest-method-to-read-maxfield-from-a-qvd/

This thread might also be useful

Where Exists() and resident load

Not applicable
Author

I didnt explain myself correctly - I do load the tables from QVD but I dont have maximum value or date value

and the ID column isn't continuous.

Therefore i can't use the "Delta load" /"Incremental load"

Is there any other way,

Anil_Babu_Samineni

How about this?

Qualify *;

Table1:

Load ID, Fname, Lname from Table1.qvd (qvd);

Unqualify ID;

Concatenate

Qualify *;

Table2:

Load ID, Fname, Lname from Table2.qvd (qvd);

Unqualify ID;


Final_Table:

Noconcatenate

Load * Resident Table1 where not exists (Table2.ID);

Best Anil, When applicable please mark the correct/appropriate replies as "solution" (you can mark up to 3 "solutions". Please LIKE threads if the provided solution is helpful
adamdavi3s
Master
Master

Oh sorry I didn't mean to imply the incremental load would be needed at all, just that loading from QVD with a single 'exists' item is very quick indeed.

So your script would look something really straightforward like like:

Fact:

Load *

from qvd1(qvd);

Load *

from QVD2(qvd)

WHERE NOT EXISTS (ID);


That will just give you one big table with all the data in, AFAIK based on what you have told us

Not applicable
Author

I tried to build like this, I thought there might be a quicker method,

(i load millions of records)

adamdavi3s
Master
Master

Nope via QVD is by the fastest! we load 120million and it takes about 12 seconds *edit* sorry no it is 79seconds */edit* to load and validate (from memory I could be wrong)

Note that if you're using more than one field in the exists then it will be slower.