ipv6 hangs in CentOS7, how to fix it?

Question

CentOS

There is a server with CentOS (esxi virtual machine), on the main interface eno16777984 looking outside, 2 IPs are configured, one ipv4, the second ipv6. The essence of the problem is that ipv6 regularly falls off, after restarting the network it rises, works no longer than a day and again goes to sleep. IPv4 on this interface works fine, without a single break :)

cat /etc/sysconfig/network
# Created by anaconda
NETWORKING_IPV6=yes
NETWORKING=yes

cat /etc/sysconfig/network-scripts/ifcfg-eno16777984 
TYPE="Ethernet"
BOOTPROTO="dhcp"
DEFROUTE="yes"
PEERDNS="yes"
PEERROUTES="yes"
IPV4_FAILURE_FATAL="no"
IPV6INIT="yes"
IPV6_AUTOCONF="yes"
IPV6_DEFROUTE="yes"
IPV6_PEERDNS="yes"
IPV6_PEERROUTES="yes"
IPV6_FAILURE_FATAL="no"
NAME="eno16777984"
UUID="9d2ac7bd-c1d8-492f-a1c8-c3b3d591b679"
DEVICE="eno16777984"
ONBOOT="yes"
IPV6ADDR="2a01:4f8:173:1413::2/64"
IPV6_DEFAULTGW="fe80::1"

cat /etc/redhat-release 
CentOS Linux release 7.3.1611 (Core)

eno16777984: flags=4163<UP,BROADCAST,RUNNING,MULTICAST>  mtu 1500
        inet xxx.xxx.xxx.xx  netmask 255.255.255.192  broadcast xxx.xxx.xxx.xx
        inet6 fe80::250:56ff:fe00:b043  prefixlen 64  scopeid 0x20<link>
        inet6 2a01:4f8:173:1413::2  prefixlen 64  scopeid 0x0<global>
        ether 00:50:56:00:b0:43  txqueuelen 1000  (Ethernet)
        RX packets 323398861  bytes 176485185521 (164.3 GiB)
        RX errors 0  dropped 0  overruns 0  frame 0
        TX packets 289485959  bytes 410142676014 (381.9 GiB)
        TX errors 0  dropped 0 overruns 0  carrier 0  collisions 0