Ticket ID: SIXXS #12883825 Ticket Status: User PoP: ruled01 - EDPnet N.V. (St.Petersburg)
Tunnel randomly stops delivering packets
Shadow Hawkins on Wednesday, 24 December 2014 16:05:01
Hello,
Once every few weeks I receive an email about my tunnels downtime. The tunnel is T121674. Last time this happened on Dec 23 but only today I could look into the issue. So it seems that my side of the tunnel was not resonding to pings since around Dec 21. But Im pretty sure that my side of tunnel is up and no configurations changes were made for at least a month or so.
I used my IPv6-capable VPS from the other part of the Earth (blua.kir.elagin.me) to ping the gateways endpoint (2a02:578:5002:c7::1) and it was responding.
Then I started pinging my tunnel endpoint (2a02:578:5002:c7::2) and it was NOT responding. Still pinging I started `tcpdump -i sixxs` on my host to check packets going through the virtual interface and there were no packets.
Then I pinged the gateways endpoint from the host that is my endpoint. And here is what I saw in tcpdump:
~~~~~
tcpdump: WARNING: sixxs: no IPv4 address assigned
error : ret -1
tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
listening on sixxs, link-type RAW (Raw IP), capture size 65535 bytes
#
# Silence until I pinged the tunnel from my side
#
18:41:57.130295 IP6 cl-200.led-01.ru.sixxs.net > gw-200.led-01.ru.sixxs.net: ICMP6, echo request, seq 1, length 64
18:41:57.131545 IP6 gw-200.led-01.ru.sixxs.net > cl-200.led-01.ru.sixxs.net: ICMP6, echo reply, seq 1, length 64
18:41:57.319875 IP6 blua.kir.elagin.me > cl-200.led-01.ru.sixxs.net: ICMP6, echo request, seq 109, length 64
18:41:58.132042 IP6 cl-200.led-01.ru.sixxs.net > gw-200.led-01.ru.sixxs.net: ICMP6, echo request, seq 2, length 64
18:41:58.133024 IP6 gw-200.led-01.ru.sixxs.net > cl-200.led-01.ru.sixxs.net: ICMP6, echo reply, seq 2, length 64
18:41:58.320479 IP6 blua.kir.elagin.me > cl-200.led-01.ru.sixxs.net: ICMP6, echo request, seq 110, length 64
18:41:59.133481 IP6 cl-200.led-01.ru.sixxs.net > gw-200.led-01.ru.sixxs.net: ICMP6, echo request, seq 3, length 64
18:41:59.134585 IP6 gw-200.led-01.ru.sixxs.net > cl-200.led-01.ru.sixxs.net: ICMP6, echo reply, seq 3, length 64
18:41:59.321640 IP6 blua.kir.elagin.me > cl-200.led-01.ru.sixxs.net: ICMP6, echo request, seq 111, length 64
# and so on
~~~~~
As you can see exactly the moment I pinged the gateway from my side I started receiving echo requests from my remote VPS.
So the question is: whats going on? Is it some kind of misconfiguration on my side or on the gateways side?
~~~~~
# uname -a
Linux verda.kir.elagin.me 3.9.5-hardenedverda #2 SMP Thu Nov 27 15:36:10 MSK 2014 x86_64 Intel(R) Core(TM)2 CPU 6300 @ 1.86GHz GenuineIntel GNU/Linux
# ip a s sixxs
8: sixxs: <POINTOPOINT,NOARP,UP,LOWER_UP> mtu 1280 qdisc noqueue state UNKNOWN
link/sit 10.188.37.24 peer 77.109.111.178
inet6 2a02:578:5002:c7::2/64 scope global
valid_lft forever preferred_lft forever
inet6 fe80::abc:2518/128 scope link
valid_lft forever preferred_lft forever
~~~~~
State change: user
Jeroen Massar on Wednesday, 24 December 2014 16:36:05
The state of this ticket has been changed to user
Tunnel randomly stops delivering packets
Jeroen Massar on Wednesday, 24 December 2014 16:37:26
Please use the forums for resolving your problem. This is not a problem on the SixXS side, hence we cannot resolve your problem for you.
Most very likely it is related to connection tracking and similar issues.
link/sit 10.188.37.24 peer 77.109.111.178
Definitely a connection tracking issue.
Posting is only allowed when you are logged in. |