I'm declaring that this is the bug for 3.5. This has been merged into ocp and is in OCP v3.5.0.37 or newer.
Verify on openshift v3.5.0.37. restart master works well and we don't meet this bugs after one days testing. [root@ip-172-18-8-15 ~]# openshift version openshift v3.5.0.37 kubernetes v1.5.2+43a9be4 etcd 3.1.0
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2017:0884