How to properly manage/ document reload_structure? - no Publisher

[datanibbler]

Hi,

we have no Publisher here, only the QlikView_server and our reload_structure is fast becoming unintuitive. There is a multitude of apps to be loaded in different intervals - from a weekly schedule to a 10min-schedule in one case - and some are dependent on the data provided by others.

Is there some "best practice" method to document all this? Mind that it's not quite sufficient to just represent the as-is_state - we need to have some possibility of taking notes rgd. dependencies between apps so we can at once see all the consequences when someone requests his app to be reloaded at another time...

Thanks a lot!

Best regards,

DataNibbler

[deepakqlikview_123]

Hi,

If you do not have publisher licemce then using server u can reload all QVWs.

You have to first create task and schedule that as per requirement.

Thanks,

Deepak

[ashfaq_haseeb]

Hi,

I would suggest to make a note of all your schedule tasks in excel.

If you have some tasks which needs to be run upon completion of another, Then make sure you run dependent task.

Make sure to note schedule time and approximate end time.

By this you will have control on what tasks run at what time and if you wan't to run any random task, then you will know its dependency.

Also use qvd Dictionary form rob to get more such details with your qvd usage.

Regards

ASHFAQ

[datanibbler]

Hi Ashfaq,

yes, Excel was m.o.l. what I was thinking about - but this requires some thinking before we actually start doing it.

That QVD_Dictionary will be very helpful, however, to find out dependencies - which qvd_files are used by which apps, which will tell us what apps have to run first. Then we can take those apps'  approximate run_time and we'll know what is the earliest time when we can run any other apps depending on those.

[ashfaq_haseeb]

Exactly.

You might need to reschedule all task by splitting them.

As you say it required lot of effort.

Regards

ASHFAQ

[Bill_Britt]

HI,

You may not want to hear this, but it might be time for you to think about purchasing Publisher.

Bill

[schlettig]

If you have an enterprise publisher you can make dependencies on more than one successful reload of a document.

When you have for instance:

RL_QVD_Doc1

RL_Datamodel_Doc1

RL_QVW_Doc1

RL_QVD_Doc2

RL_Datamodel_Doc2

RL_QVW_Doc2

RL_QVW_Doc3

And Doc3 can only start after successful reload of QVW_Doc1, QVW_Doc2 than you can put in such a trigger and don't need to calculate the run time.

BR,

Christian

[datanibbler]

Hi Bill,

I don't mind in the least. Only, it's not my decision - and there is absolutely no chance of this company purchasing Publisher at the current price - even less right now than at any other time, "we" are currently on a saving_trip.

The problems are in fact elsewhere, but in the eyes of a manager, cutting costs always looks sexy because you can present something to your superiors right away ...

(I never said this, this post will self-destroy in 30sec)

[datanibbler]

Hi,

I have thought about the possibilities of the Publisher more than once - but well, we can only work with the reality that we have.

First things first: Is there any way I can export the status_quo (when what app is scheduled for reload now) out of the QMC?

That would give me a basis. Then I would just need to add the average reload_time and the dependencies.

Best regards,

DataNibbler

[schlettig]

Hi,

you would find the configuration and logs of the reloads in C:\ProgramData\QlikTech\DistributionService in a default installation.

You could use the QlikView System Monitor to help you calculate average Load Times:

http://community.qlik.com/docs/DOC-4307

BR,

Christian