I wanted to know if HE could add the ability to adjust the tunnels MTU through the web interface from the default of 1280 up to whatever your particular link type is capable of handling?
What sort of link do you have?
Quote from: eonesixfour on April 08, 2008, 07:02:16 PM
What sort of link do you have?
It is irrelevant what I have. To be safe the default is set lower than anything you would encounter whether it is a system co-lo'd somewhere, DSL, cable and so on. I would like to be able to adjust the MTU to match the capability of the circuit type.
We'll consider the suggestion.
Quote from: brad on April 08, 2008, 08:50:10 PM
It is irrelevant what I have.
I fail to see how, you might have an internet2 link which has a rather large MTU, then again if you are using a HE tunnel this isn't very likely. There is a very good reason for the MTU being set to what it is, since if you have a cable connection or PPPoA you'll most likely have a MTU of 1500, PPPoE is 1492 etc, to run IPv6 over IPv4 you have a bunch of overheads and bumping the MTU will just fragment packets possibly unnecessarily etc.
I'm getting an MTU of 1514 on my T0 interface
Quote from: rlhdomain on April 11, 2008, 02:06:11 PM
an MTU of 1500 is fine all the way to 100Mbps
so unless you need jumbo frames then don't worry
The MTU of the *tunnel* is what matters, not the physical interface or virtual interface (PPPoE) providing your Internet connection. This also has nothing to do with Jumbo frames.
Quote from: eonesixfour on April 09, 2008, 11:56:40 PM
I fail to see how, you might have an internet2 link which has a rather large MTU, then again if you are using a HE tunnel this isn't very likely. There is a very good reason for the MTU being set to what it is, since if you have a cable connection or PPPoA you'll most likely have a MTU of 1500, PPPoE is 1492 etc, to run IPv6 over IPv4 you have a bunch of overheads and bumping the MTU will just fragment packets possibly unnecessarily etc.
I am not talking about the MTU of my connection. I am talking about the MTU of the *tunnel*. The default and minimum MTU of the tunnel is 1280. Well below any Ethernet, cable, PPPoE, etc connection. This lower MTU will cause unnecessary fragmentation. The tunnel has an overhead of 20 bytes for the IPv4 header meaning the maximum MTU of the tunnel can be is 1480 for a connection with an MTU of 1500.
Quote from: rlhdomain on April 12, 2008, 11:36:53 PM
have you even looked at the set bandwidth of the tunnel?
The bandwidth of a virtual interface (VLAN/Tunnel) is for all intents and purposes irrelevant.
Although I have used tons of Cisco gear I am not using Cisco gear at the moment for these tunnels.
Quote from: rlhdomain on April 12, 2008, 11:36:53 PM
the MTU only needs to be raised if the bandwidth is high and so is the quality
This is not true and even if it were true I have no issue with bandwidth or quality
of the bandwidth on my end.
Quote from: rlhdomain on April 12, 2008, 11:36:53 PM
on my cisco router if I type "show interface tunnel0" then I see the Bandwidth thats set and the MTU also your argument of fragmentation is invalid as a lower MTU gives less fragmentation and is more useful at lower bandwidth
Yes there is a threshold where it should be smaller than the smallest MTU of the connections between my router and HE's router but it should also be higher than the default which is currently being used.
If I was on a modem or ISDN connection then I would agree but bandwidth is not an issue.
Quote from: rlhdomain on April 12, 2008, 11:36:53 PM
MTU= Maximum transmition unit= Max size for per packet in bytes= 1.5 kilobytes (or more accurately 1.46KB)
and although your wan interface may be able to do 1500 without fragmentation it may not be able to do a tunnel with 1500
I never said a tunnel with an MTU of 1500. I said 1480 or in the case of a PPPoE connection 1472.
Quote from: rlhdomain on April 12, 2008, 11:36:53 PM
since the tunnels have differing MTU's I would assume its automaticly set by the HE point based on what can work with minimal fragmentation
The tunnels with HE do not have differing MTU's at the moment.
OP: This sounds as if it's a problem with your specific configuration. My system reports a default mtu of 1480 as expected where the real interface is 1500. I don't see how you're getting a default of 1280. Please explain your hardware configuration. Something is different.
I can't help if you can't understand
I wanted a response from Hurricane Electric personnel involved with the Tunnel broker service, not from people who have no idea what they're talking about or random opinions.
FYI we are still looking into it as a configuration option inside the broker. There are just other things that require our attention at the moment.
Quote from: broquea on April 25, 2008, 10:03:54 PM
FYI we are still looking into it as a configuration option inside the broker. There are just other things that require our attention at the moment.
Great, thank you very much.
Quote from: snarked on April 13, 2008, 04:50:20 PM
OP: This sounds as if it's a problem with your specific configuration. My system reports a default mtu of 1480 as expected where the real interface is 1500. I don't see how you're getting a default of 1280. Please explain your hardware configuration. Something is different.
I just checked mine too, and I'm seeing the MTU automatically being set at 1480 as well.
Quote from: broquea on April 25, 2008, 10:03:54 PM
FYI we are still looking into it as a configuration option inside the broker. There are just other things that require our attention at the moment.
Any word on this? The ability to set a higher MTU could help with transfer rates and overall performance. Count me among the interested. Thanks in advance!
personally, i think this is a non-issue, mtu should be automatically negotiated to the highest possible, meaning that of the link with the lowest. I suspect brad's problem may be due to a certain howto, http://wiki.debian.org/DebianIPv6 (http://wiki.debian.org/DebianIPv6), which lists:
Quoteauto 6in4
iface 6in4 inet6 v4tunnel
address [Your IPv6 Endpoint]
netmask [Prefix Length]
endpoint [PoP IPv4 Endpoint]
ttl 64
up ip link set mtu 1280 dev 6in4
up ip route add default via [PoP IPv6 Endpoint] dev 6in4
... or something similar
I'm curious whether that's just a typo, or if there is a reason for it...
but what do i know, i'm just a n00b ;D
Quote from: malfeasant on December 28, 2008, 07:43:12 AM
personally, i think this is a non-issue, mtu should be automatically negotiated to the highest possible, meaning that of the link with the lowest. I suspect brad's problem may be due to a certain howto, http://wiki.debian.org/DebianIPv6 (http://wiki.debian.org/DebianIPv6), which lists:
Quoteauto 6in4
iface 6in4 inet6 v4tunnel
address [Your IPv6 Endpoint]
netmask [Prefix Length]
endpoint [PoP IPv4 Endpoint]
ttl 64
up ip link set mtu 1280 dev 6in4
up ip route add default via [PoP IPv6 Endpoint] dev 6in4
... or something similar
I'm curious whether that's just a typo, or if there is a reason for it...
but what do i know, i'm just a n00b ;D
Should be except that is not how tunnels work, and I don't even run Debian never mind a Linux-based OS.
I tried setting the MTU of the tunnel interface on my router to 1480, but whenever I do a PMTU discovery over it the tunnel server drops it back down to 1280. MTUs are not automatically negotiated with 6in4 tunneling and must be set manually on both sides. Incidentally, 1280 is the smallest allowed IPv6 packet according to RFC 2460, but a smaller MTU means fragmentation is more likely and can hinder performance.
Does HE support "Internet2" size MTUs in its network? Please note that I'm NOT limiting my question to those using a tunnel server (since I also have NATIVE IPv6 with HE as the peering carrier). I noted that here in Los Angeles, HE peers with cenic.net (which uses Internet 2 MTUs). Or is 1,500 bytes a limit in HE's transit network?
Transit peering link:
10gigabitethernet2-2.core1.lax1.he.net (2001:470:1:1b::1)
lax-hpr--he-peer.cenic.net (2001:468:e00:801::1)
(extracted from a traceroute6).
1500 isn't a transit network limitation, we use jumbo frames wherever possible. Most of the links in LA, for example are 9212.
Thanks for the info. Now I have to figure out how to do it via an ethernet interface (server to router).
Just found out from my co-lo facility that they can't get me onto a gigabit port within their internal network so I can reach HE's feed into there via a jumbo frame. :-( However, I still have a dedicated 100baseT IPv6-only ethernet cable/port connection. Their Cisco router should handle it, but it's the intervening router that's the problem. (However, when they upgrade, I'll be ready to go.)
Any harm leaving my MTU at 9216? OK, so I'll get one ICMP packet back (from the internal network) and switch the MTU to 1500, but why not just leave it?
Quote from: r1ske on November 15, 2008, 07:03:27 PM
Quote from: broquea on April 25, 2008, 10:03:54 PM
FYI we are still looking into it as a configuration option inside the broker. There are just other things that require our attention at the moment.
Any word on this? The ability to set a higher MTU could help with transfer rates and overall performance. Count me among the interested. Thanks in advance!
<AOL>Me too!</AOL> :)
Any word? Are you still on it?
I have a troublesome configuration: an Apple Time Capsule (or Airport) loses packets with IP payloads >1496 bytes. That's 1496 +8 PPPoE = 1504 over-large Ethernet packet payload. Sadly, BT, my ISP, insists on shoving large frames down peoples' throats, operating at 1500-byte MTU and relying on either jumbo frames or MSS clamping to make the connection useful. So until I find hardware that supports baby jumbos >1504 payload, I'd need HE to let me tune the tunnel MTU to 1476.
But of course, you already know that. :)
Cheers,
Sabahattin
PS: In theory, this should all be self-regulating, since HE can respond to IPv4 ICMP "Packet too big and DF set MTU=x" messages. Not in this case, however.