IPv6 unreachable until restart interface

IPv6 works just well in most time, but everyday 's midnight, our network is cut up by school to force us go to bed, and recovered at next day 's 6am. This is done, in my guess, by justing power off the dorm's switch.

And after the cutting up in the next day morning, IPv4 works still fine, but my IPv6 cannot connect to any host except roommates. If I restart the interface by /etc/init.d/net.enp4s0 restart, then everything works again.

Here is the topology:

         My machine \        
                     room's switch - dorm's switch - dorm's gateway 
Roommate 's machine /    

This is what it looks like before restart:

Riaqn-Laptop ~ # ip -6 route
2001:da8:8000:e0b1::/64 dev enp4s0  proto kernel  metric 256  expires 3584sec
fe80::/64 dev enp0s29f7u2u4  proto kernel  metric 256 
fe80::/64 dev enp4s0  proto kernel  metric 256 
default via fe80::212:cfff:fe5f:c0c0 dev enp4s0  proto ra  metric 1024  expires 1784sec
Riaqn-Laptop ~ # ip -6 neigh
2001:da8:8000:e0b1::15f7 dev enp4s0  FAILED
2001:da8:8000:e0b1::3b4e:1660 dev enp4s0  FAILED
fe80::f928:6537:a988:7b80 dev enp4s0  FAILED
2001:da8:8000:e0b1::fe71 dev enp4s0  FAILED
2001:da8:8000:e0b1::a113 dev enp4s0  FAILED
2001:da8:8000:e0b1::58aa dev enp4s0 lladdr b8:88:e3:8f:30:7e STALE
fe80::212:cfff:fe5f:c0c0 dev enp4s0 lladdr 00:12:cf:5f:c0:c0 router STALE
2001:da8:8000:e0b1::15f8 dev enp4s0  FAILED
2001:da8:8000:e0b1::569e dev enp4s0 lladdr dc:0e:a1:72:9e:8b STALE
fe80::c9fe:6d0e:9295:75a dev enp4s0 lladdr dc:0e:a1:72:9e:8b STALE
Riaqn-Laptop ~ # ping6 -I enp4s0 fe80::212:cfff:fe5f:c0c0
PING fe80::212:cfff:fe5f:c0c0(fe80::212:cfff:fe5f:c0c0) from fe80::21b:24ff:fed5:f438 enp4s0: 56 data bytes
^C
--- fe80::212:cfff:fe5f:c0c0 ping statistics ---
6 packets transmitted, 0 received, 100% packet loss, time 5001ms
Riaqn-Laptop ~ # ip -6 addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 
    inet6 ::1/128 scope host 
       valid_lft forever preferred_lft forever
2: enp4s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qlen 1000
    inet6 2001:da8:8000:e0b1::cdf4/64 scope global 
       valid_lft forever preferred_lft forever
    inet6 fe80::21b:24ff:fed5:f438/64 scope link 
       valid_lft forever preferred_lft forever
4: enp0s29f7u2u4: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qlen 1000
    inet6 fe80::8eae:4bff:fe00:4e82/64 scope link 
       valid_lft forever preferred_lft forever
Riaqn-Laptop ~ # rdisc6 enp4s0
Soliciting ff02::2 (ff02::2) on enp4s0...
Timed out.
Timed out.
Timed out.
No response.

This is how looks like after restarting interface

Riaqn-Laptop ~ # ip -6 addr
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 
    inet6 ::1/128 scope host 
       valid_lft forever preferred_lft forever
2: enp4s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qlen 1000
    inet6 2001:da8:8000:e0b1::9b59/64 scope global 
       valid_lft forever preferred_lft forever
    inet6 fe80::21b:24ff:fed5:f438/64 scope link 
       valid_lft forever preferred_lft forever
4: enp0s29f7u2u4: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qlen 1000
    inet6 fe80::8eae:4bff:fe00:4e82/64 scope link 
       valid_lft forever preferred_lft forever
Riaqn-Laptop ~ # ip -6 route
2001:da8:8000:e0b1::/64 dev enp4s0  proto kernel  metric 256  expires 3599sec
fe80::/64 dev enp0s29f7u2u4  proto kernel  metric 256 
fe80::/64 dev enp4s0  proto kernel  metric 256 
default via fe80::212:cfff:fe5f:c0c0 dev enp4s0  proto ra  metric 1024  expires 1799sec
Riaqn-Laptop ~ # ip -6 neigh
fe80::212:cfff:fe5f:c0c0 dev enp4s0 lladdr 00:12:cf:5f:c0:c0 router REACHABLE
Riaqn-Laptop ~ # rdisc6 enp4s0
Soliciting ff02::2 (ff02::2) on enp4s0...

Hop limit                 :           64 (      0x40)
Stateful address conf.    :          Yes
Stateful other conf.      :          Yes
Mobile home agent         :           No
Router preference         :       medium
Neighbor discovery proxy  :           No
Router lifetime           :         1800 (0x00000708) seconds
Reachable time            :  unspecified (0x00000000)
Retransmit time           :  unspecified (0x00000000)
 Source link-layer address: 00:12:CF:5F:C0:C0
 Prefix                   : 2001:da8:8000:e0b1::/64
  On-link                 :          Yes
  Autonomous address conf.:           No
  Valid time              :         3600 (0x00000e10) seconds
  Pref. time              :         1800 (0x00000708) seconds
 from fe80::212:cfff:fe5f:c0c0
^C

The IPv6 address is assigned by DHCPv6, this is how dhclient work:

Riaqn-Laptop ~ # dhclient -6 -v   -e IF_METRIC=2 -1 -pf /var/run/dhclient-enp4s0.pid enp4s0
Internet Systems Consortium DHCP Client 4.3.1 Gentoo-r0
Copyright 2004-2014 Internet Systems Consortium.
All rights reserved.
For info, please visit https://www.isc.org/software/dhcp/

Waiting 5 seconds for IPv6 to really come ready
Listening on Socket/enp4s0
Sending on   Socket/enp4s0
Created duid \000\001\000\001\0343\234\205\000\033$\325\3648.
PRC: Soliciting for leases (INIT).
XMT: Forming Solicit, 0 ms elapsed.
XMT:  X-- IA_NA 24:d5:f4:38
XMT:  | X-- Request renew in  +3600
XMT:  | X-- Request rebind in +5400
XMT: Solicit on enp4s0, interval 1050ms.
RCV: Advertise message on enp4s0 from fe80::212:cfff:fe5f:c0c0.
RCV:  X-- Preference 10.
RCV:  X-- IA_NA 24:d5:f4:38
RCV:  | X-- starts 1419829253
RCV:  | X-- t1 - renew  +345600
RCV:  | X-- t2 - rebind +552960
RCV:  | X-- [Options]
RCV:  | | X-- IAADDR 2001:da8:8000:e0b1::b5b0
RCV:  | | | X-- Preferred lifetime 691200.
RCV:  | | | X-- Max lifetime 1036800.
RCV:  X-- Server ID: 00:01:00:00:53:7c:9e:c4:00:15:5d:7f:03:00
RCV:  Advertisement recorded.
PRC: Selecting best advertised lease.
PRC: Considering best lease.
PRC:  X-- Initial candidate 00:01:00:00:53:7c:9e:c4:00:15:5d:7f:03:00 (s: 154, p: 10).
XMT: Forming Request, 0 ms elapsed.
XMT:  X-- IA_NA 24:d5:f4:38
XMT:  | X-- Requested renew  +3600
XMT:  | X-- Requested rebind +5400
XMT:  | | X-- IAADDR 2001:da8:8000:e0b1::b5b0
XMT:  | | | X-- Preferred lifetime +7200
XMT:  | | | X-- Max lifetime +7500
XMT:  V IA_NA appended.
XMT: Request on enp4s0, interval 1000ms.
RCV: Reply message on enp4s0 from fe80::212:cfff:fe5f:c0c0.
RCV:  X-- IA_NA 24:d5:f4:38
RCV:  | X-- starts 1419829254
RCV:  | X-- t1 - renew  +345600
RCV:  | X-- t2 - rebind +552960
RCV:  | X-- [Options]
RCV:  | | X-- IAADDR 2001:da8:8000:e0b1::b5b0
RCV:  | | | X-- Preferred lifetime 691200.
RCV:  | | | X-- Max lifetime 1036800.
RCV:  X-- Server ID: 00:01:00:00:53:7c:9e:c4:00:15:5d:7f:03:00
PRC: Bound to lease 00:01:00:00:53:7c:9e:c4:00:15:5d:7f:03:00.

[Update]

I just found that the problem may be unrelated to the network cutting up. This morning I restarted the interface to provide the above information, and now, after 9 hours, it's unreachable already without being cutting up, and in a more interesting way: I can ping ipv6.google.com, but cannot ping the router.

Riaqn-Laptop ~ # ping6 ipv6.google.com
PING ipv6.google.com(hkg03s16-in-x08.1e100.net) 56 data bytes
64 bytes from hkg03s16-in-x08.1e100.net: icmp_seq=1 ttl=43 time=340 ms
64 bytes from hkg03s16-in-x08.1e100.net: icmp_seq=2 ttl=43 time=340 ms
64 bytes from hkg03s16-in-x08.1e100.net: icmp_seq=3 ttl=43 time=339 ms
^C
--- ipv6.google.com ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2002ms
rtt min/avg/max/mdev = 339.998/340.070/340.141/0.675 ms
Riaqn-Laptop ~ # ping6 -I enp4s0 fe80::212:cfff:fe5f:c0c0
PING fe80::212:cfff:fe5f:c0c0(fe80::212:cfff:fe5f:c0c0) from fe80::21b:24ff:fed5:f438 enp4s0: 56 data bytes
^C
--- fe80::212:cfff:fe5f:c0c0 ping statistics ---
4 packets transmitted, 0 received, 100% packet loss, time 3007ms

Riaqn-Laptop ~ # ndisc6 fe80::212:cfff:fe5f:c0c0 enp4s0        
Soliciting fe80::212:cfff:fe5f:c0c0 (fe80::212:cfff:fe5f:c0c0) on enp4s0...
Timed out.
Timed out.
Timed out.
No response.
Riaqn-Laptop ~ # ip -6 neigh
fe80::212:cfff:fe5f:c0c0 dev enp4s0 lladdr 00:12:cf:5f:c0:c0 router REACHABLE
fe80::c9fe:6d0e:9295:75a dev enp4s0  FAILED
Riaqn-Laptop ~ # ping6 -I enp4s0 fe80::212:cfff:fe5f:c0c0
PING fe80::212:cfff:fe5f:c0c0(fe80::212:cfff:fe5f:c0c0) from fe80::21b:24ff:fed5:f438 enp4s0: 56 data bytes
^C
--- fe80::212:cfff:fe5f:c0c0 ping statistics ---
5 packets transmitted, 0 received, 100% packet loss, time 4000ms

Riaqn-Laptop ~ # ip -6 route
2001:da8:8000:e0b1::/64 dev enp4s0  proto kernel  metric 256 
fe80::/64 dev enp0s29f7u2u4  proto kernel  metric 256 
fe80::/64 dev enp4s0  proto kernel  metric 256 
default via fe80::212:cfff:fe5f:c0c0 dev enp4s0  proto ra  metric 1024  expires 1764sec

Ok, I asked a friend using windows 7 on his computer, to keep his machine on from midnight to the morning, and yes, his IPv6 is broken too.

So I guess this is an upstream problem, and cannot be fixed on my own. Fortunately they provide an isatap tunnel to let me connect to IPv6 through a IPv4 tunnel, which works fine even after several hours.

I think my problem solved, thanks for your help!