SixXS::Sunset 2017-06-06

Ticket ID: SIXXS #12116461
Ticket Status: User

PoP: nlams05 - SURFnet (Amsterdam)

High Latency Loss
[nl] Shadow Hawkins on Friday, 08 August 2014 09:35:49
I have two SixXS tunnels, T6437 and T126970. Since a month or so, T126970 is showing around 40% Latency Loss here on the SixXS status page. A screenshot I took a little few weeks ago: http://tweakers.net/ext/f/QzJksjaP4RN5tVrObcnAN3oL/full.png The other tunnel is showing completely healthy. Both are static 6in4 Both use the same PoP: nlams05 Both are on the same ISP, Ziggo, but in different towns Both are terminated on Asus RT-AC66U routers running Merlin firmware with similar settings T6437 is terminated on a router behind a bridged Ubee docsis modem T126970 is terminated on a router in the DMZ of a Ubee docsis modemrouter No changes have been made on my side around the time the loss started. Unless my ISP changed something that affects only 1 of my two tunnels. In practice both tunnels are working fine. Connectivity is fast and bandwidth also seems fine. Running ping commands to the PoP IPv4 or to IPv6 hosts on the internet show no loss and ~10ms response times. Nevertheless, I am curious what could be causing this latency loss and whether I could (and should?) do something about it? The only thing I tried a week ago is tinkering with the MTU settings of T126970. This used to be on the default 1280. I increased it to 1480, tunnel is still working fine, but no change to the latency loss reported by SixXS.
State change: user Locked
[ch] Jeroen Massar SixXS Staff on Friday, 08 August 2014 09:50:04
Message is Locked
The state of this ticket has been changed to user
High Latency Loss
[ch] Jeroen Massar SixXS Staff on Friday, 08 August 2014 09:59:12
Please use the forums for resolving your problems. The ticket system is for when there is something that is wrong on the SixXS side. See amongst others: FAQ: Tunnel endpoint didn't ping and .
Both are terminated on Asus RT-AC66U routers running Merlin firmware with similar settings
Similar, thus not the same?
Running ping commands to the PoP IPv4 or to IPv6 hosts on the internet show no loss and ~10ms response times.
That works as that is outbound traffic, likely due to connection tracking inbound packets do not work though as then there is no state.
The only thing I tried a week ago is tinkering with the MTU settings of T126970.
The pings are smaller than 1280 as if they where not they would have to be fragmented/split-up for most tunnels, hence latency check packets are not bothered with MTU.
High Latency Loss
[nl] Shadow Hawkins on Friday, 08 August 2014 16:26:40
Thanks, and my apologies for using the ticket system rather than the forums. I was hoping to get some feedback from SixXS on what the cause of this latency loss could be. The main difference is that the site that has the problems has a modemrouter (NAT) in between, so perhaps this device started dropping 40% of the latency packets somehow.
High Latency Loss
[ch] Jeroen Massar SixXS Staff on Friday, 08 August 2014 16:39:42
the site that has the problems has a modemrouter (NAT) in between,
NAT == connection tracker. See the FAQ mentioned above for how that can give problems.

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

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