I would suggest to go with dotTrace to see if you can spot anything obvious. it could be that one of the entries has circular reference.
you're welcome to contact developer support service for help. feel free to mention me and I'm happy to take a look
Thanks for your suggestions.
Currently I'm investigating if this is the issue:
This is our current Catalog structure:
Catalog
-> RootCategory
-> Category 1
-> Category 11
-> Category 12
-> Category 13
-> Category 14
-> Category 2
-> Category 21
-> Category 22
-> Category 23
-> Category xxxx...
It seems like when you move or delete anything below "RootCategory" then Optimizely starts loop everything.
Could that be the issue?
Definitely something with the catalog structure. deletion is one of the test cases and I'm sure we can't get it release if it does not pass QA (i.e takes too long to delete)
Turns out is was Find that was the issue. Temporary disabling EventIndexing for Find, while moving the node, solved the issue.
Hi Daniel, whenever you create or delete anything find indexing kicks, it could be because of that.
Yes, a bit strange that it locks the thread, I thought it should be handle in the background by another thread. The Find index is not that important in this case so I just switch the listener off while doing move or delete on nodes and then turn it back on again.
We have an integration that handles the Commerce catalog structure.
When we try to move Catalog nodes within the structure it takes really long time to process (moving within the same catalog). Even if the node only have 4 child entries it still takes really long time (+10 min).
I have tried with contentRepository.Move, relationRepository.SetParentNode and the old classig catalogSystem.Save, same result with all.
Even if i try to move the node manually within commerce UI it still takes forever (the UI just keeps showing loading icon).
I investigated the debug logs for Episerver Commerce and it looks like it keeps looping entries endlessly.
I have tried disabling all catalog event listeners.
Any ideas on how to proceed with troubleshooting, or should I simply accept that moving nodes takes really long time?