Skip to main content
Announcements
Join us at Qlik Connect for 3 magical days of learning, networking,and inspiration! REGISTER TODAY and save!
cancel
Showing results for 
Search instead for 
Did you mean: 
Anonymous
Not applicable

Storing of QVD's for Historical Data

Hi All,

Scenario:

We have a pre-existing extract QVW that runs at intervals of 15mins 24/7 but this extract only pulls the last 14 days worth of data.

This is done in this way as a result of the sheer size of the files and performance of the server, as currently the performance is a bit compromised.

There is often times a requirement from management to see more than 14days worth of data for analytical purposes.  So what I would like to accomplish is to keep a historical QVD file on my personal PC and when needed present this to management.

My question is two-fold:

lets say I copy the QVD from the server every few days, obviously the QVD's will then include the majority of the same information.

If i were to load all the QVD's into a QVW and use the store function (which will then obviously concatenate the tables), would this create duplicate values?

Furthermore, ideally, i would like to run a script on the server itself that can automatically save this data periodically, lets say at 12 at night everyday.  How would I ensure that no duplicates are stored (keep in mind that i do have a Timestamp field in the data) so I imagine loading distinct would be adequate, ,however, the file would exceed 20million rows of data within a week and consist of about 30 columns of data, thus my concern is really the performance thereof.

The table mentioned above I will not include for the purpose of the question, instead I will give an example of a much smaller tables fields.

Script:

LOAD MANDT,

          LGBEL,

          TASKS

          AREA,

          QUEUE,

          WHO,

          CONFIRMATION_TIME

FROM

[..\..\QVLive\ORDIM_C.QVD]

(qvd);

Can I ask for some assistance with this regard?

Below is an example of what I had in mind, my concern is just that the variable will look at the MAX(CONFIRMATION_TIME) of the final concatenated table instead of the initial loaded table, which would result in the table not adding any new data.

My proposed solution:

LET vMaxdate = max(CONFIRMED_AT);

STOREDQVD:

LOAD

MANDT,

          LGBEL,

          TASKS

          AREA,

          QUEUE,

          WHO,

          CONFIRMATION_TIME

FROM

[..\..\QVData\QVLive\StoredQVD.QVD]

(qvd);

ORIGINALQVD:

LOAD

          MANDT,

          LGBEL,

          TASKS

          AREA,

          QUEUE,

          WHO,

          CONFIRMATION_TIME

FROM

[..\..\QVData\QVLive\OriginalQVD.QVD]

(qvd)

where CONFIRMATION_TIME>$(vMaxdate);

STORE STOREQVD INTO StoredQVD(qvd);

1 Reply
rahulpawarb
Specialist III
Specialist III

Hello Ruan,

To generate QVDs you can create task at QMC level with said time intervals; you can also break the file into parts based on certain field value such as one QVD per year or per region etc. Post that you can load these files into your QlikView applications. This will help you on performance front.

Please go through the attached pdf file to get more details on Incremental Load; This will definitely help you.

Post that Let me know if you have any queries.

P.S.: Thanks to Authors of these two docs.

Regards!

Rahul