NPrinting

[eric67]

Bonjour,

Je suis en train de migrer une application Nprinting 16.1.0.0 vers la version 19.29.3.0

Je rencontre 2 soucis majeurs, et 1 plus mineur :

Le premier : je constate que lorsque je lance une tâche de publication, avec plusieurs rapports à envoyer par mail, il génère autant de fois chaque rapport qu'il y a d'adresses email destinataires.  Une tâche qui prenait quelques dizaines de secondes dans l'ancienne version, prend désormais plusieurs minutes (voir dizaines de minutes...)

Le deuxième : dans l'ancienne version, il était possible de créer des jobs, ce qui permettait de lancer les tâches les unes à la suite des autres. Aujourd'hui, sauf erreur d'interprétation de ma part, il n'y a plus de job, et sauf à mettre des horaires de lancement pour chacune des tâches (dans les déclencheurs), il est impossible de lancer toutes les tâches à la suite, car en les sélectionnant, et en cliquant sur exécuter, toutes les tâches se lancent en même temps, et quand il en a des centaines, le serveur se plante... Et je ne trouve pas de paramétrage pour déclarer une "file d'attente maxi".

Et le troisième plus mineur : il n'y a plus la fonction de duplication d'une tâche, c'était bien commode lorsque d'une tâche à l'autre il n'y avait que quelques filtres à modifier.

Si vous avez des éléments de réponses à m'apporter, je suis preneur.

Merci d'avance

Bien cordialement

Eric

[Lech_Miszkiewicz]

Hi Eric,

I strongly suggest to use "English" here as you will get much larger chance to get an answer:

Here is your post translated using "Google Translate":

Hello,

I am in the process of migrating an Nprinting 16.1.0.0 application to version 19.29.3.0

I have 2 major concerns, and 1 more minor:

The first: I notice that when I launch a publication task, with several reports to send by email, it generates each report as many times as there are recipient email addresses. A task that took a few tens of seconds in the old version now takes several minutes (see tens of minutes ...)

The second: in the old version, it was possible to create jobs, which made it possible to launch the tasks one after the other. Today, unless there is an interpretation error on my part, there are no more jobs, and except for putting launch times for each of the tasks (in the triggers), it is impossible to launch all the tasks at the same time. continuation, because by selecting them, and by clicking on execute, all the tasks are launched at the same time, and when there are hundreds of them, the server crashes ... And I cannot find any configuration to declare a "file d 'maximum wait'.

And the third more minor: there is no longer the function of duplicating a task, it was very convenient when from one task to another there were only a few filters to modify.

If you have any answers to give me, I am interested.

thank you in advance

Best regards

Eric

Before I start I would like to ask if you have gone through any NPrinting training? It is enterprise class tool and often people assume it is easy to set things up as it is just "Drag & Drop", but this is nothing more than big mistake. You need proper planning, thinking about solution from perspective and simple 1:1 migration will not work! 

Answers:

1. This is a case if recipients have user filters and in such scenario reports must be created by NPrinting individually as those filters can be different for each of them. Typically if there are no user filters for report you would put this is report in dedicated NPrinting app, bunch up users into NPrinitng group and setup group as recipient. That would be my approach and in such scenario you would not get duplicated reports. Other options is to use April 2020 or newer NPrinting and setup one dummy user for the report with variable containg emails of all users who will receive it. more here: https://nprintingadventures.com/2020/06/17/small-big-change-dynamic-email-address/

2. You are correct - there are no jobs but at the same time to set it up correctly you need to learn little bit about NPrinting first and consider environments. The fact that your server crashes leads me to think that you are using "QlikView - Local" connection instead of QVP. QVP would be recommended connection to allow for stability and best performance! When you say: "all the tasks are launched at the same time, and when there are hundreds of them, the server crashes ..." what do you mean - which server crashes, QlikView or NPrinitng? As I mentioned below even if you trigger 100 tasks but you run NPrinting Engine on server say with 4 CPU cores, only 4 QV.exe session will be open at the time and all other tasks will wait in the queue anyway. Designing this requires knowledge of workload but also understanding which server does heavy lifting when using LOCAL vs QVP connection.

At the same time you should study documentation on https://help.qlik.com and learn about CPU/RAM recommendations for QlikView reports. Number of CPU core will dictate number of parallel QlikView sessions open by NPrinting. You cannot compare this to version 16 as it was not multi-threaded. NPrinitng 17+ is multi-threaded and allows for multiple reports generation at a time. 

If you really require task dependencies you can use API and task chaining method by setting it up in QMC. Details here: https://nprintingadventures.com/2019/04/08/nprinting-api-qlik-rest-subroutines/

3. I agree with you - it is dearly missed feature!

hope this helps

regards

Lech

[Lech_Miszkiewicz]

....one last question - why do you migrate to such old version. I strongly suggest to migrate to September 2020 which is 6-7 releases newer (including patches) and resolves tonnes of bugs which were present in older versions. It also has a lot more new features and functionalities.