For anyone reaching here - The problem was that the logstsash.db file was stuck and somehow corrupted. I deleted it, and then it was automatically created again correctly and logs started flowing.

For anyone debugging this kind of issue:

There was no error indication on this problem in my application logs or in the different ELK components logs. All the logs I looked at had no errors or warnings. Only after following all the pipeline carefully and understanding the flow I learned about this file that is used for synchronization and I noticed in my environment it was not being updated at all. Sadly, I have lost all the logs from the gap between the time it stopped and until I deleted it. I think there probably is some way to re-scan the files with the data to index them (but I will leave this issue for a different question)