Rene Viskar

True, did not see that as a possibility. We have `entity_id` fields in our entity table so the fix works well for us at the moment. But as to solving...

@romainruaud Thanks for asking. Seems to be exactly the issue we had. I've solved it as stated above, as it works for us very well and we had limited resources...

@romainruaud our service provider will upgrade our staging and production servers to MariaDb 10.4 in a few days time. So we will find out soon enough, if the newer version...

Hi! Can confirm, exactly the same problem occurring here, with identical flow and exactly the same versions. @duffner any chance you had any luck figuring this out? I've also narrowed...

The local web hosting service we use is responsible for hosting the elasticsearch service, but take it as a self-hosting Elastic as it is installed on the same private server.

@vahonc thanks for investigating. Did you try it in a multi-store environment? For us it still happens regularly (about once a week, because of our custom product importer cron), but...

@vahonc that's what we did, but for reproducing purposes it allowed us to kind-of reproduce the issue. The problem still stands even with scheduling, that it seems like the indexer...