[Extended Sync] How to model parallel access?

[anschl]

We currently read data from the EDC like this:

1. The REST client checks if there is an EDR for the asset
2. The client creates the EDR if none was found
3. The client calls /aas/request and returns the payload

This has a chance of creating the EDR twice if two requests come in with an unlucky timing, leading to an error in step 3. Do I need to take care of that or is there a solution for this in the EDC?

[anschl]

Summary of the answer I got on the call:
Conflicts are possible, however it should only be a problem while testing, as creating multiple EDRs rapidly is not expected in production use. EDRs can be compared to opening a socket connection in this regard.

[florianrusch-zf]

Hi, could you or someone quickly explain why it's not expected, that in production rapidly multiple EDRs will be created?

[anschl]

I will leave the answer to someone more knowledgeable than I am. In case you need the answer in a reasonable time, I recommend joining the weekly call and ask there.