[c-nsp] Cisco 2600 with MPLS, what of MTU?
James Weathersby (jweather)
jweather at cisco.com
Wed May 31 07:47:29 EDT 2006
MPLS is not supported on the 2600s except the 2691. Most of the MPLS
subsystems have been removed from the images. There's no way to
increase the MTU, so any larger packets will be fragmented or dropped.
james
James Weathersby, TME
Access Technology Group
Cisco Systems
7025 Kit Creek Rd.
RTP, NC 27709
Phone 919-392-0984
Mobile 919-931-0137
-----Original Message-----
From: cisco-nsp-bounces at puck.nether.net
[mailto:cisco-nsp-bounces at puck.nether.net] On Behalf Of
cisco-nsp-request at puck.nether.net
Sent: Wednesday, May 31, 2006 7:16 AM
To: cisco-nsp at puck.nether.net
Subject: cisco-nsp Digest, Vol 42, Issue 81
Send cisco-nsp mailing list submissions to
cisco-nsp at puck.nether.net
To subscribe or unsubscribe via the World Wide Web, visit
https://puck.nether.net/mailman/listinfo/cisco-nsp
or, via email, send a message with subject or body 'help' to
cisco-nsp-request at puck.nether.net
You can reach the person managing the list at
cisco-nsp-owner at puck.nether.net
When replying, please edit your Subject line so it is more specific than
"Re: Contents of cisco-nsp digest..."
Today's Topics:
1. Re: Serial interface flapping/Automatic shutdown (Kevin Graham)
2. Re: 6500s/XFPs (Was XENPAK 10GB-LW (WAN PHY))
(Asbjorn Hojmark - Lists)
3. Re: Serial interface flapping/Automatic shutdown (Wolfgang Roth)
4. Re: 6500s/XFPs (Was XENPAK 10GB-LW (WAN PHY)) (Mikael Abrahamsson)
5. Re: 6500s/XFPs (Was XENPAK 10GB-LW (WAN PHY))
(Richard A Steenbergen)
6. Re: Cisco & Juniper POS interoperability (Diego Castro)
7. Re: 6500s/XFPs (Was XENPAK 10GB-LW (WAN PHY))
(Richard A Steenbergen)
8. Cisco 2600 with MPLS, what of MTU? (Kristian Larsson)
9. Re: 6500s/XFPs (Was XENPAK 10GB-LW (WAN PHY))
(Asbjorn Hojmark - Lists)
10. Re: 6500s/XFPs (Was XENPAK 10GB-LW (WAN PHY))
(Asbjorn Hojmark - Lists)
11. Re: 6500s/XFPs (Was XENPAK 10GB-LW (WAN PHY)) (Marian Durkovic)
12. SIP and NAT/PAT issues (Church, Chuck)
----------------------------------------------------------------------
Message: 1
Date: Tue, 30 May 2006 13:28:09 -0700
From: "Kevin Graham" <mahargk at gmail.com>
Subject: Re: [c-nsp] Serial interface flapping/Automatic shutdown
To: "Wolfgang Roth" <Wolfgang.Roth at brave.de>
Cc: cisco-nsp at puck.nether.net
Message-ID:
<2a64fada0605301328n3b48513fk3e1e48a030686766 at mail.gmail.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
On 5/30/06, Wolfgang Roth <Wolfgang.Roth at brave.de> wrote:
> I ask, because if an interface flaps rapidly (say every few seconds)
> and you use a routing protocol like OSPF the load on the involved
> routers increases drastic and the network becomes some kind of
unstable.
Check out 'IP Event Dampening':
http://www.cisco.com/univercd/cc/td/doc/product/software/ios120/120newft
/120limit/120s/120s22/s_ipevdp.htm
------------------------------
Message: 2
Date: Tue, 30 May 2006 22:55:13 +0200
From: "Asbjorn Hojmark - Lists" <lists at hojmark.org>
Subject: Re: [c-nsp] 6500s/XFPs (Was XENPAK 10GB-LW (WAN PHY))
To: "'Ed Butler'" <ed.butler at rapidswitch.com>
Cc: cisco-nsp at puck.nether.net
Message-ID: <001201c6842b$59102bd0$280a0a0a at hojmark.net>
Content-Type: text/plain; charset="us-ascii"
> Do Cisco have anything in the pipeline for XFPs on 6500s?
Well there's the XFP-10GLR-OC192SR in SPA-1XTENGE-XFP in 7600- SIP-600,
but...
> (That will be comparable in cost to the 6704/Xenpak solution)
>From what I've been told, no. The argument was that multimode is
very important for a lot of their big (enterprise?) customers and that
you can't do LX4 on XFPs.
Coming modules should be X2-based, however.
-A
------------------------------
Message: 3
Date: Tue, 30 May 2006 23:08:34 +0200
From: Wolfgang Roth <Wolfgang.Roth at brave.de>
Subject: Re: [c-nsp] Serial interface flapping/Automatic shutdown
To: Tassos Chatzithomaoglou <achatz at forthnet.gr>, Kevin Graham
<mahargk at gmail.com>
Cc: cisco-nsp at puck.nether.net
Message-ID: <7B9363C3A6AA3215E7286B93@[10.0.0.2]>
Content-Type: text/plain; charset=us-ascii; format=flowed
> Also have a look at event dampening on newer IOS:
> Check out 'IP Event Dampening':
Great, thank you - exactly what I looked for!
Do you have any recommondations for the configurable parameters when
using low bandwidth (E1) serial interfaces? The default values sound OK
for me, but probably you can do a graduation about this subject...
Wolfgang
------------------------------
Message: 4
Date: Tue, 30 May 2006 23:19:02 +0200 (CEST)
From: Mikael Abrahamsson <swmike at swm.pp.se>
Subject: Re: [c-nsp] 6500s/XFPs (Was XENPAK 10GB-LW (WAN PHY))
To: cisco-nsp at puck.nether.net
Message-ID: <Pine.LNX.4.62.0605302317480.20178 at uplift.swm.pp.se>
Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed
On Tue, 30 May 2006, Asbjorn Hojmark - Lists wrote:
>> From what I've been told, no. The argument was that multimode is
> very important for a lot of their big (enterprise?) customers and that
> you can't do LX4 on XFPs.
>
> Coming modules should be X2-based, however.
EW. X2 is Xenpak in another form factor and most likely has the same
pricepoint.
I also haven't seen anyone but Cisco 4500 BU use X2 so that doesn't bode
well for the price to go down either.
--
Mikael Abrahamsson email: swmike at swm.pp.se
------------------------------
Message: 5
Date: Tue, 30 May 2006 19:22:46 -0400
From: Richard A Steenbergen <ras at e-gerbil.net>
Subject: Re: [c-nsp] 6500s/XFPs (Was XENPAK 10GB-LW (WAN PHY))
To: Asbjorn Hojmark - Lists <lists at hojmark.org>
Cc: cisco-nsp at puck.nether.net
Message-ID: <20060530232246.GM19934 at overlord.e-gerbil.net>
Content-Type: text/plain; charset=us-ascii
On Tue, May 30, 2006 at 10:55:13PM +0200, Asbjorn Hojmark - Lists wrote:
> > Do Cisco have anything in the pipeline for XFPs on 6500s?
>
> Well there's the XFP-10GLR-OC192SR in SPA-1XTENGE-XFP in 7600-
> SIP-600, but...
>
> > (That will be comparable in cost to the 6704/Xenpak solution)
>
> >From what I've been told, no. The argument was that multimode is
> very important for a lot of their big (enterprise?) customers and that
> you can't do LX4 on XFPs.
>
> Coming modules should be X2-based, however.
X2 is pretty much the worst of all worlds. The tradeoff with XENPAK vs
XFP is that you get first crack at the new high-end optics which require
more space to implement in the first gen. For example, DWDM ITU grid
tuned XENPAKs with 80km reach (ZRD) have been shipping for quite a
while, but they're just now starting to trickle into the market for XFP.
X2 is still XAUI based (its reall just XENPAK in a smaller form factor),
but with a significantly less competition^Wexisting users. I've never
seen a ZR X2 in the wild let alone a ZRD or WAN PHY. The only things
switching to X2 are going to get you is an even bigger headache in
sparing, a requirement to buy new pluggables, and higher prices across
the board.
Don't get me wrong, keeping a XENPAK based board around for legacy
optics like LX4 and for early adoption of new long reach optics makes
sense, but when the high density 10GE cards start rolling out they
should really be XFP. Otherwise Cisco is going to get its ass kicked by
folks like Foundry and Force10 (F10 24-port 10GE XFP in 1U for $22k list
- SWEET!) in the dense/cheap 10GE market.
--
Richard A Steenbergen <ras at e-gerbil.net>
http://www.e-gerbil.net/ras
GPG Key ID: 0xF8B12CBC (7535 7F59 8204 ED1F CC1C 53AF 4C41 5ECA F8B1
2CBC)
------------------------------
Message: 6
Date: Tue, 30 May 2006 16:23:30 -0700 (PDT)
From: Diego Castro <dcastro5 at yahoo.com>
Subject: Re: [c-nsp] Cisco & Juniper POS interoperability
To: cisco-nsp at puck.nether.net
Message-ID: <20060530232330.7225.qmail at web32403.mail.mud.yahoo.com>
Content-Type: text/plain; charset=iso-8859-1
Thanks to everyone, this issue has been solved.
The problem was the IOS version.
The first trial was with 12.0(31)S1 service provider, and the protocol
never came up.
In the lab we tried with 12.0(31)S5 service provider, and everything was
working.
This is the Juniper M10 configuration:
interfaces {
}
so-1/1/0 {
encapsulation cisco-hdlc;
sonet-options {
fcs 32;
payload-scrambler;
rfc-2615;
}
unit 0 {
family inet {
no-redirects;
address 10.1.0.249/30;
}
}
}
chassis {
fpc 1 {
pic 1 {
framing sdh;
}
}
}
This is the Cisco 7206VXR configuration:
interface POS1/0
ip address 10.1.0.250 255.255.255.252
crc 32
pos framing sdh
pos scramble-atm
saludos
Diego
--- Diego Castro <dcastro5 at yahoo.com> wrote:
>
> nishal
>
> Yes, we have the same configuration:
> pos framing sdh
> pos scramble-atm
>
> This router was working very good with version 12.1(26)E4, but the new
> card requiere a differente IOS 12.0(31)S1.
> After reboot with the hardware and IOS 12.0(31)S1, the old POS
> interface doesn't like to bring up the protocol.
> If a restarted again with the old IOS, everything is fine.
>
> Tomorrow I will implement a lab with my Cisco and the Juniper box to
> find out was is missing or changed.
> I will inform all of you.
>
> diego
> --- nishal goburdhan <nishal at is.co.za> wrote:
>
> > On Sat, May 27, 2006 at 10:36:18AM -0700, Diego Castro wrote:
> > > Mark
> > >
> > > I had a 7206VXR running 12.1(26)E4 service
> > provider
> > > using PA-POS-OC3SMI or PA-POS that is connected
> to
> > my
> > > ISP that has a Juniper M10.
> > > Now we got a new card PA-POS-2OC3 that requiere
> a
> > new
> > > version and is recommended 12.0(30)S1.
> > > I installed that version on my router but after restarted the POS
> > > interfaces doesn't show up/up,
> > only
> > > the protocol was down.
> > > The new card is not installed, because I need
> that
> > > both cards be in the same router.
> > > The first step was to install the new software
> > with
> > > the old hardware without luck.
> > > So i was wondering that maybe is needed some
> > change in
> > > the Juniper or Cisco configuration.
> >
> > it's possible that the code change changed some of the default
> > settings.
> > have you confirmed with your ISP that you've both got the same
> > framing set (sdh/sonet)
> > - pos framing sdh
> >
> > are you both using scrambing (or not)
> > - pos scramble-atm
> >
> >
> > 'sh contr pos x/y' usually has some useful
> debugging
> > info...
> >
> >
> > --n.
> >
>
> _______________________________________________
> cisco-nsp mailing list cisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at
> http://puck.nether.net/pipermail/cisco-nsp/
>
------------------------------
Message: 7
Date: Tue, 30 May 2006 19:46:17 -0400
From: Richard A Steenbergen <ras at e-gerbil.net>
Subject: Re: [c-nsp] 6500s/XFPs (Was XENPAK 10GB-LW (WAN PHY))
To: Mikael Abrahamsson <swmike at swm.pp.se>
Cc: cisco-nsp at puck.nether.net
Message-ID: <20060530234617.GN19934 at overlord.e-gerbil.net>
Content-Type: text/plain; charset=us-ascii
On Tue, May 30, 2006 at 11:19:02PM +0200, Mikael Abrahamsson wrote:
>
> EW. X2 is Xenpak in another form factor and most likely has the same
> pricepoint.
>
> I also haven't seen anyone but Cisco 4500 BU use X2 so that doesn't
> bode well for the price to go down either.
Well, a bunch of folks I know are using the HP 6410cl-6xg 6-port X2
switches and swear by them, but the future of 10GE networking they
certainly aren't. The only other place they make sense is stuffing them
into legacy or stackable systems which can't support XFP. Cisco sells
the
X2 modules it does have at the exact same price as XENPAK at least.
The real problem is I'm not aware of anyone even making ZR/ZRD/LW/etc in
X2 form, let alone anyone selling them to end users. This means there is
no way serious networking using X2 could be contemplated in anything
under
1 year. Even if they started developing and selling all of the missing
optics, by the time they roll out XFPs are going to be signifiantly
cheaper and more dense. In the WAN PHY department scaling XENPAK/X2 is
pretty much hopeless, there is only so much you can do to make them
price efficient compared to XFPs. Personally I think you'd have to be
crazy to invest in X2 infrastructure at this point.
BTW today the list price difference between LR and LW xenpaks is $10k
($2k vs $12k respectively). Compare this to being able to use the same
XFP to support both, at around the $2k mark (or less), where the cost of
the framer is theoretically shifted from the pluggable to the host. Of
course, essentially every high end (not counting some of the lower end
stackables which happen to have an XFP port) 10GE XFP based platform
today supports LAN/WAN PHY already, either by default or for a
difference per card that is in the $1k/ea or less range.
--
Richard A Steenbergen <ras at e-gerbil.net>
http://www.e-gerbil.net/ras
GPG Key ID: 0xF8B12CBC (7535 7F59 8204 ED1F CC1C 53AF 4C41 5ECA F8B1
2CBC)
------------------------------
Message: 8
Date: Wed, 31 May 2006 07:06:58 +0200
From: Kristian Larsson <kristian at spritelink.se>
Subject: [c-nsp] Cisco 2600 with MPLS, what of MTU?
To: cisco-nsp at puck.nether.net
Message-ID: <20060531050658.GC3242 at spritelink.se>
Content-Type: text/plain; charset=us-ascii
It would seem MPLS is supported on the 2600 platform but as the mpls
seemingly increases your packet size with a few bytes will the ethernet
controllers on the 2600 be able to cope with this or will the resulting
MTU be lower?
Kristian.
------------------------------
Message: 9
Date: Wed, 31 May 2006 09:19:47 +0200
From: "Asbjorn Hojmark - Lists" <lists at hojmark.org>
Subject: Re: [c-nsp] 6500s/XFPs (Was XENPAK 10GB-LW (WAN PHY))
To: "'Richard A Steenbergen'" <ras at e-gerbil.net>, "'Mikael
Abrahamsson'" <swmike at swm.pp.se>
Cc: cisco-nsp at puck.nether.net
Message-ID: <001f01c68482$98eef590$280a0a0a at hojmark.net>
Content-Type: text/plain; charset="us-ascii"
> Cisco sells the X2 modules it does have at the exact same price as
> XENPAK at least.
Yes, 4 k$ for the 10Gbase-LR... which just happens to be the same as for
the XFP-10GLR-OC192SR. Well, at least it's consistent.
-A
------------------------------
Message: 10
Date: Wed, 31 May 2006 09:29:37 +0200
From: "Asbjorn Hojmark - Lists" <lists at hojmark.org>
Subject: Re: [c-nsp] 6500s/XFPs (Was XENPAK 10GB-LW (WAN PHY))
To: "'Richard A Steenbergen'" <ras at e-gerbil.net>
Cc: cisco-nsp at puck.nether.net
Message-ID: <002101c68483$f8e00150$280a0a0a at hojmark.net>
Content-Type: text/plain; charset="us-ascii"
>> Coming modules should be X2-based, however.
> X2 is pretty much the worst of all worlds.
I believe it's just about the only way you can put eight ports on a C6K
blade and still support LX4. Since they have customers who insist on
multimode, the choice must have been pretty easy.
-A
------------------------------
Message: 11
Date: Wed, 31 May 2006 10:01:49 +0200
From: Marian Durkovic <md at bts.sk>
Subject: Re: [c-nsp] 6500s/XFPs (Was XENPAK 10GB-LW (WAN PHY))
To: Richard A Steenbergen <ras at e-gerbil.net>
Cc: cisco-nsp at puck.nether.net
Message-ID: <20060531080149.GA48232 at us.svf.stuba.sk>
Content-Type: text/plain; charset=us-ascii
On Tue, May 30, 2006 at 07:46:17PM -0400, Richard A Steenbergen wrote:
> The real problem is I'm not aware of anyone even making ZR/ZRD/LW/etc
> in
> X2 form, let alone anyone selling them to end users. This means there
> is no way serious networking using X2 could be contemplated in
> anything under
> 1 year.
Well, seeing the market directions, it might even happen that noone will
ever start developing/producing them...
M.
------------------------------
Message: 12
Date: Wed, 31 May 2006 06:16:02 -0500
From: "Church, Chuck" <cchurch at netcogov.com>
Subject: [c-nsp] SIP and NAT/PAT issues
To: "nsp" <cisco-nsp at puck.nether.net>
Message-ID:
<B6621ED4D0AD394BBA73CA657DFD8976D0D9D5 at MSPEXBE01.wamnet.inc>
Content-Type: text/plain; charset="us-ascii"
Anyone,
I'm looking for a good whitepaper on SIP call setup, and how the
RTP is setup after the SIP portion is done. I looked around a bit, and
didn't see anything with as much detail as I'm looking for. I've got a
2650 with a T1 internet connection, default route. It's doing QoS,
prioritizing VoIP in both directions. It's also doing NAT/PAT and CBAC
firewalling. Yesterday I had an issue with one way voice on a couple
different SIP phones, and an ATA. All three are different brands, and
access a SIP proxy held at a commercial VoIP provider out on the 'net.
>From what I could see from the 'sh ip inspect sess' was that CBAC got
really confused, and was trying to send an RTP stream inbound to the
wrong device. The destination it was trying didn't exist, and the ARP
entry for it was incomplete. Rebooting the router made the problem go
away. Anyone got any tips on combining PAT/SIP/CBAC? I did see some
bugs relating to 12.3.18 and one-way voice. I threw 12.4.8 on last
night, hoping it helps. The problem has happened only twice in the last
8 weeks, and rebooting the router fixed it both times, so I'm leaning
more towards a software bug versus a config issue, but any tips would be
greatly appreciated.
Thanks in advance,
Chuck Church
------------------------------
_______________________________________________
cisco-nsp mailing list
cisco-nsp at puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
End of cisco-nsp Digest, Vol 42, Issue 81
*****************************************
More information about the cisco-nsp
mailing list