NPrinting Long User Import Time

[justindallasverizon]

Hello Folks,

I have an NPrinting system that takes a while (~12 hours) to import all the users.  What should my next move be to shorten this process?  I know I can get a larger servers, but are there transaction settings on the DB that I might be able to look into?

Any help is greatly appreciated.

The import will also fail, complaining about a deadlock if that helps.

[Lech_Miszkiewicz]

Hi @justindallasverizon 

That is just crazy - something is either wrong or you really are loading such large number of users that it crashes whole thing!

• What is the number of users you are trying to load?
• What process are you using?
  • native excel?
  • native LDAP?
    • are you using LDAP filters so that only required users are loaded.
  • other: API load?
• Are there other task running in parallel when you try to do that?
• Are there any unfinished tasks running in background which should be killed?
• Are you using NewsStand for your distribution causing your repository to grow largely
  • if yes - have you checked its size?
• Do you have enough disk space on C drive - I saw very unstable environments due to lack of space on harddrive

It would also be important to understand number of filters you are potentially loading and overall architecture of your server. 

I am assuming you are meeting the bare minimum requirements

• (2 cores for server + 2 cores for engine  = 4 CPU cores
• minimum 8 GB of RAM which I assume you would have 16GB anyway given that we live in year 2025 and half of it is consumed by OS

Share your specs if you can and also workload of the server: number of tasks running + number of recipients. Just rough estimate) and we will take it from there.

cheers

Lech

[justindallasverizon]

1. What is the number of users you are trying to load?

120,000 users, 17,000 filters

2. What process are you using?

Excel Import

3. Are there other task running in parallel when you try to do that?

No

4. Are there any unfinished tasks running in background which should be killed?

No

5. Are you using NewsStand for your distribution causing your repository to grow largely

No

6. Do you have enough disk space on C drive - I saw very unstable environments due to lack of space on harddrive

We have 166GB of free space on the server

Thank you for your quick response.  Let me know if you need more information.  Also, do you think another engine would help?

[hugo_andrade]

Hi @justindallasverizon ,

Thanks for bringing all information @Lech_Miszkiewicz  requested. That's a huge help.

Would you be able to also share a screenshot of the task configuration?

Due to the large amount of users, I would recommend you splitting the task into multiple ones until you reach a successful run.

We can help you with that.

[Lech_Miszkiewicz]

What are the hardware specs of the server?

having this large number of users requires large server and frankly i am jot even sure if single server will handle that number of recipients. 

@Ruggero_Piccoli - can you comment on that?

[justindallasverizon]

Here is a pic of the task config.  Do you think getting another engine would help?

[hugo_andrade]

That's a smart question. But another engine wouldn't help on this case. The import task is executed by the NPrinting Scheduler, not the NPrinting Engine.

So, I assume this is an LDAP Import task.

Are you able to reduce the scope of the task by breaking it into different groups?

On the Advanced section, you can add LDAP filter for users. I'm giving you an example to load all users that are member of the group "GroupName". You will need to use the group Distinguished name for it.

(&(objectCategory=Person)(sAMAccountName=*)(memberOf:1.2.840.113556.1.4.1941:=cn=GroupName,cn=Users,dc=Company,dc=local))

Then, you create another task for another group. The caution is that all users should be covered by the join of all the groups you list.

[hugo_andrade]

@justindallasverizon ,

It might be valuable to monitor the disk usage, CPU and RAM usage on the server during the 12h window.

[Lech_Miszkiewicz]

HI @hugo_andrade 

as you can see from previous answers it is not LDAP, it is excel import. You can also see from documentation that large deployments are considered when you are loading say 700 users. https://help.qlik.com/en-US/nprinting/February2025/Content/NPrinting/DeployingQVNprinting/Multi-engi... .

In this case we have 170x larger volume so it is rather a question to Qlik ( @Ruggero_Piccoli @Frank_S ) if they have designed it to support such large number of users at all?

@justindallasverizon - I have asked about your server hardware details. What can you tell about that? I am sure you realized that loading such large number of users creates quite a complex relationship in repository so it may be rather important to have this information.

cheers

Lech

[Frank_S]

checking...

cc @Ruggero_Piccoli