SixXS::Sunset 2017-06-06

Ticket ID: SIXXS #645190
Ticket Status: User

PoP: 

TunnelRobot falsely reporting my Tunnel as being down
[us] Shadow Hawkins on Monday, 31 December 2007 03:43:54
Tag [/b] is not closed
State change: user Locked
[ch] Jeroen Massar SixXS Staff on Monday, 31 December 2007 03:58:16
Message is Locked
The state of this ticket has been changed to user
TunnelRobot falsely reporting my Tunnel as being down
[ch] Jeroen Massar SixXS Staff on Monday, 31 December 2007 03:59:03
TunnelRobot falsely reporting my Tunnel as being down
[us] Shadow Hawkins on Monday, 31 December 2007 21:14:00
I've read the FAQ, as you so elgently tried to imply for me to read via email. What you've quoted is not my issue. My tunnel state is not the issue, nor is my stateful inspection, logs never indicate my tunnel state as fluctuating. I'm running a Cisco 1841 router, this is not an issue with ipchains. I've had this same ipv6 configuration in my router since I first came up, but only recenly been getting notified by your tunnelrobot that there's an issue, even though every time I look, I see my router responding to icmps via ipv6 and the next hop responding to ipv6 icmps, even during the time your tunnel robot says I wasn't. I've got a co-worker setup through sixxs.net on the same pop, and he's been getting emails recently also, and your mrtg graphs show us both not responding to ICMPs at the same time. I do not believe this to be an issue with my router's configuration, but I'll begin logging ipv6 icmp's over time.
TunnelRobot falsely reporting my Tunnel as being down
[ch] Jeroen Massar SixXS Staff on Tuesday, 01 January 2008 10:28:49
[..]logs never indicate my tunnel state as fluctuating.
Which logs are supposedly telling you that proto-41 tunnels can 'fluctuate'? There is no state in the proto-41 itself, they are just IPv6 inside IPv4 packets, nothing more, nothing less; as such there is no way that the protocol itself can tell if the tunnel is up or not (if there was we would be using it). That is why we use ICMPv6 from ::1 to ::2, and clearly your endpoint didn't respond.
I'm running a Cisco 1841 router, this is not an issue with ipchains.
It does not matter what implementation of "Connection Tracking" you have, if you have connection tracking on your host and it is tracking proto-41 packets, the state for this will time out and thus incoming packets will be rejected.
[..] your mrtg graphs show us [..]
Which "mrtg graphs"? Nowhere in SixXS is mrtg used to monitor anything.
TunnelRobot falsely reporting my Tunnel as being down
[us] Shadow Hawkins on Tuesday, 01 January 2008 23:13:15
The log i'm referring to is the general "show log" in the cisco ios. I assume if the tunnel peer was not reachable for whatever reason, line protocol would have shown down like it has in the past when my ISP was on the fritz. It's possible there's an issue with my stateful inspection, but it's kinda odd that this would have just crept up as a problem now as I've been running this config since I went live with you, and I cannot find any references to a bug tracking protocol 41 packets on Cisco's site. I'll span a port over on my switch on my public side and start a sniffer capture and let it run for a few weeks and see if it happens again. If the issue is my router, it will be quite obvious. The graph I was referring to are under the Tunnel statistics, I just assumed you guys were using MRTG as they look identical to graphs generated by MRTG.
TunnelRobot falsely reporting my Tunnel as being down
[ch] Jeroen Massar SixXS Staff on Wednesday, 02 January 2008 10:13:08
As said, proto-41 doesn't have any idea of 'line up/down', as such your logs won't contain it either. It is not MRTG as MRTG is not usable for our purposes, it does use RRD though.

Please note Posting is only allowed when you are logged in.

Static Sunset Edition of SixXS
©2001-2017 SixXS - IPv6 Deployment & Tunnel Broker