The 2nd one doesn't seem to affect anything, but the first one seems correlated to these logs for failed requests happening in production