Seems like both your connections experience this lack of response from KSE at the same time but one is configured to allow more keep alive failures than the other.
So, maybe you can tweak those settings, but also there might be something about your network or environment causing this failure rather than KSE lagging or not responding for some reason, assuming those are 2 different servers.