Re: UDP encaps for SCTP and SCCP

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On May 22, 2010, at 10:18 PM, <L.Wood@xxxxxxxxxxxx> <L.Wood@xxxxxxxxxxxx> wrote:

> Randall,
> 
> If you're referring to RBSCP, which is a weak TCP/SCTP PEP that
> uses SCTP between tunnel endpoints, then yes, that is available
> in SCTP-capable Cisco routers. But it's not in widespread use.
> 
> having a large number of IANA PPIDs for SCTP is simply
> an indication that SCTP is overspecified. (Does anyone
I do not understand why having PPIDs means overspecified.
When writing a dissector for SCTP it is very handy to have an
indication which upper layer protocol is encapsulated in a
DATA chunk. If an upper layer doesn't want to do it, he can
always use 0.
> even use partial reliability?)
Yes. And it will hit the IETF eventually...

Best regards
Michael
> 
> 
> On 22 May 2010, at 13:43, Randy Stewart wrote:
> 
>> There are several projects I have worked on that
>> DO use SCTP internally. Do they run on the big I.. in a way,
>> they are in routers.
>> 
>> Are they behind NAT's no.. since most NAT's with two exceptions do
>> not support SCTP.
>> 
>> There are other protocols that I am aware of in Datacenters as well that
>> have picked SCTP.
>> 
>> If you look at the IANA registry for Protocol ID's in SCTP I think there
>> are 28 PPID's assigned... now thats not a huge number but its also
>> not representative since many folks just pick their own PPID and use  
>> it.. not
>> paying attention to the registry.
>> 
>> 
>> R
>> On Apr 27, 2010, at 12:27 AM, <L.Wood@xxxxxxxxxxxx>  
>> <L.Wood@xxxxxxxxxxxx> wrote:
>> 
>>> I don't think that counts as a we-chose-to-use-SCTP end user...
>>> 
>>> SIGTRAN infrastructure in the telephony providers doesn't go through  
>>> NAT. ergo, SCTP's most popular use doesn't need to go through NAT.  
>>> And DCCP doesn't need to go through NAT, because no-one uses DCCP...  
>>> so there's no need to tunnel these over UDP to get through NAT.
>>> 
>>> actual applications running on an actual end-user computer with a  
>>> TCP/IP stack, please.
>>> 
>>> On 26 Apr 2010, at 23:49, Brian F. G. Bidulock wrote:
>>> 
>>>> L.Wood,
>>>> 
>>>> Hmmm.  SIGTRAN requires SCTP.  SIGTRAN is widely
>>>> deployed and running your phone calls.
>>>> 
>>>> --brian
>>>> 
>>>> L.Wood@xxxxxxxxxxxx wrote:          (Mon, 26 Apr 2010 21:34:14)
>>>> 
>>>> --x-- snip --x--
>>>>> On a related note, I spent years dealing with proponents of
>>>>> SCTP, and yet never encountered any actual
>>>>> we-chose-to-use-SCTP end users. So, are there any actual DCCP
>>>>> users out there?
>>>> 
>>>> -- 
>>>> Brian F. G. Bidulock
>>>> bidulock@xxxxxxxxxxx
>>>> http://www.openss7.org/
>>> 
>>> Lloyd Wood
>>> L.Wood@xxxxxxxxxxxx
>>> http://sat-net.com/L.Wood
>>> 
>>> 
>>> 
>> 
>> -----
>> Randall Stewart
>> randall@xxxxxxxxxxxx
>> 
>> 
>> 
>> 
> 
> Lloyd Wood
> L.Wood@xxxxxxxxxxxx
> http://sat-net.com/L.Wood
> 
> 
> 
> 



[Index of Archives]     [Linux Kernel Development]     [Linux DCCP]     [IETF Annouce]     [Linux Networking]     [Git]     [Security]     [Linux Assembly]     [Bugtraq]     [Yosemite]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Linux SCSI]     [DDR & Rambus]

  Powered by Linux