[BUG] pagination does not improve response time

[mmaeffert]

Describe the bug When using the ?limit query parameter, it does not improve the response time compared to not setting a limit at all

To Reproduce Steps to reproduce the behavior:

1. Create 300k twins
2. use GET /submodels endpoint
3. see that the response takes more than a minute
4. use GET /submodels?limit=5 endpoint
5. See that the response takes equally long

Expected behavior I would expect a faster response time nearly equivalent to the set limit / number of submodels ratio

BaSyx (please complete the following information):

• BaSyx latest
• BaSyx Java SDK
• AAS Version [e.g. DotAAS V2, DotAAS V3]

Additional context I suspect that application still querries every submodel and just adds the first five to the response

[zhangzai123]

Hi @mmaeffert , thank you for your report. What kind of persistence / backend do you use? InMemory or MongoDb?

[mmaeffert]

I am using MongoDB. FYI: I hosted it locally on my laptop, I suppose hosting it remotely would increase the response time even further.

[zhangzai123]

Thank you for your fast response. We are aware of the pagination issue with MongoDB backend. We are working on this with the PR eclipse-basyx/basyx-java-server-sdk#516 . The pagination response time will be improved after this PR is merged.

[mmaeffert]

This is great news, thank you. However, is there a way to disable the functionality to get every submodel alltogether? Imagine hosting this to a customer, you would not want him to keep using that endpoint

[zhangzai123]

516