The one I linked does now have the annotation, and the trick with triggering re-indexing also seems to work. But in general, no, I still get the errors.
are you saying that if you go to a random observation that you’ve never looked at before and try to add an observation, you’ll get the same errors? if you repeat this 10 times, you’ll get the same errors 10 times?
well, i guess it’s good to learn that “server blips” can create these observations that end up in a state where the need to be reindexed. i guess this at least gives us a good event that can be investigated to figure out how to prevent or mitigate similar future problems.