Created attachment 1451558 [details] screenshot from deployment in progress
Does not seem to be a kubernetes bug, but rather a bug in the gluster client or server. Changing component to "cns-deployment" for potential tracking of RHGS bug (tbd).
> Setting needinfo on Amar to assign some gluster engineer to look into this bug. There are significant changes in RHGS 3.4.0 in rpc layer, and overall brick-mux code, which was used in CNS environments. We haven't seen any such connection issues arising in latest testing (since last 6 months or so). I am inclined to close this bug as WORKSFORME now.