Re: [c-nsp] How to bring one link down if another related link goes down

[Mikael Abrahamsson]

On Mon, 27 Sep 2010, John Neiberger wrote:

That may be the best idea I've seen. It wouldn't even be necessary for 
it to bring the links back up automatically. If one link goes down, 
we're going to need to do a lot of careful troubleshooting before we 
bring it up, so a manual restoration process would be best.


I recommend changing the metric of the routing protocol of the link 
instead of shutting it down. Set it to a very high value indicating the 
link is "a last resort". Do it at both ends.


--
Mikael Abrahamssonemail: swm...@swm.pp.se
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] How to bring one link down if another related link goes down

[Gert Doering]

Hi,

On Mon, Sep 27, 2010 at 05:12:40PM -0600, John Neiberger wrote:
> Is there some way to utilize some sort of object tracking to force one
> link to go down if its partner goes down? Alternatively, the link
> itself does not need to go down. The traffic we're dealing with here
> is statically routed, so it would suffice for the static routes to be
> removed from the routing table unless both links were up. 

It should be possible.  You can have the routers ping the interface
address of the remote for both links and you can tie static route
to the result of the ping...

I haven't tested how to setup a track object for two remote IP addresses,
so that might pose a problem.

This should give you an idea what can be done:

  
http://www.cisco.com/en/US/docs/ios/12_3/12_3x/12_3xe/feature/guide/dbackupx.html

(even if the URL says 12_3, it applies to SXH and SR* as well)

gert

-- 
USENET is *not* the non-clickable part of WWW!
   //www.muc.de/~gert/
Gert Doering - Munich, Germany g...@greenie.muc.de
fax: +49-89-35655025g...@net.informatik.tu-muenchen.de


pgp0GwGEA4291.pgp
Description: PGP signature
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] cisco-nsp Digest, Vol 94, Issue 104

[Aladi Saputra]

U
-Original Message-
From: cisco-nsp-requ...@puck.nether.net
Sender: cisco-nsp-boun...@puck.nether.net
Date: Mon, 27 Sep 2010 17:17:52 
To: 
Reply-To: cisco-nsp@puck.nether.net
Subject: cisco-nsp Digest, Vol 94, Issue 104

Send cisco-nsp mailing list submissions to
cisco-nsp@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-requ...@puck.nether.net

You can reach the person managing the list at
cisco-nsp-ow...@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: SFP info (Ingen Schenau, Jeroen van (ICTS))
   2. Nexus evolution (Seth Mattinen)
   3. Re: cisco MPLS AutoBandwidth Allocator (Peter Rathlev)
   4. Re: Nexus evolution (Chris Evans)
   5. Re: ISR G2 performance (Peter Rathlev)
   6. Re: Nexus evolution (David Freedman)
   7. Re: Nexus evolution (Tim Stevenson)
   8. Re: Nexus evolution (Quinn Snyder)
   9. Cisco 4900M BGP Support (Jimmy Changa)
  10. Re: Cisco 4900M BGP Support (Charles Klement)


--

Message: 1
Date: Mon, 27 Sep 2010 17:58:17 +0200
From: "Ingen Schenau, Jeroen van (ICTS)"

To: jack daniels 
Cc: cisco-nsp@puck.nether.net
Subject: Re: [c-nsp] SFP info
Message-ID: <1285603097.5643.178.ca...@icts-sp-039>
Content-Type: text/plain; charset="UTF-8"

> If I use SFP LX connector and coonect two routers ports back to back.
> Will this have any impact on router port .

No impact, it should work. You can connect LX transceivers back to back;
iirc, minimum patch length is somewhere around 0.5 to 2 meters.

> Is that I can use only SX SFP for Back to back connectivity with routers.

No, LX or BX are OK too. Only with ZX or other long-haul optics you
might need an attenuator.

> Also if my router port is coonecting to MUX which is colocated , I
> need to mandatorly use SX SFP.

That depends on what the MUX supports, I guess... SX is generally used
on short distances since the optics are cheaper.


Regards,

Jeroen van Ingen
ICT Service Centre
University of Twente, P.O.Box 217, 7500 AE Enschede, The



--

Message: 2
Date: Mon, 27 Sep 2010 09:32:07 -0700
From: Seth Mattinen 
To: cisco-nsp@puck.nether.net
Subject: [c-nsp] Nexus evolution
Message-ID: <4ca0c707.3030...@rollernet.us>
Content-Type: text/plain; charset=UTF-8

About a year ago there were some large-ish threads on the Nexus and a
couple people that had them in production had commented that there were
bugs that made them feel like test subjects, plus a various assortment
of unexpected limitations. How much has this changed over the last year?

I do notice that the 2248TP fabric extender supports direct to 7k, and
the 22xxTP datasheet lists 100/1000 as supported speeds. I've been
researching a 7k as a candidate for a small colo datacenter, and to me
it seems like it's matured quite a bit (on paper, anyway).

~Seth


--

Message: 3
Date: Mon, 27 Sep 2010 18:34:57 +0200
From: Peter Rathlev 
To: jack daniels 
Cc: cisco-nsp@puck.nether.net
Subject: Re: [c-nsp] cisco MPLS AutoBandwidth Allocator
Message-ID: <1285605297.11014.9.ca...@abehat.dyn.net.rm.dk>
Content-Type: text/plain; charset="UTF-8"

On Mon, 2010-09-27 at 09:21 +0530, jack daniels wrote:
> I'm stuck in the understanding - does oit make sense to implement MPLS
> AutoBandwidth in scenario where I have only 20 subnets max to be sent
> on this Backbone.

AutoBandwidth gives you the advantage of having the network recalculate
LSPs every now and then. This can (partly) overcome the scenario where
the specific time of establishing an LSP would have a negative effect on
how it's built.

If you use TE extensively and have one or more paths that are
oversubscribed (via RSVP) then AutoBandwidth may be used to shift around
some of the paths to (maybe) achieve a better utilisation.

If you don't have overlapping TE tunnels there's no point in using
AutoBandwidth. If you have plenty capacity (i.e. the sum of bandwidth of
all tunnels is less than you "narrowest" link's capacity) there's also
no need IMO.

I your network is carefully engineered off-line and you never want the
network to change you engineering you also don't need AutoBandwidth.

-- 
Peter




--

Message: 4
Date: Mon, 27 Sep 2010 12:37:44 -0400
From: Chris Evans 
To: Seth Mattinen 
Cc: cisco-nsp@puck.nether.net
Subject: Re: [c-nsp] Nexus evolution
Message-ID:

Content-Type: text/plain; charset=ISO-8859-1

We are deploying 7-5-2 like mad.

Stable platform for its age.
On Sep 27, 2010 12:34 PM, "Seth Mattinen"  wrote:
> About a year ago there were some large-ish threads on the Nexus and a
> couple people that had them in production had commented that there were
> bugs that made them 

Re: [c-nsp] Nexus evolution

[William Cooper]

I'm still a bit confused... I've a pretty significant investment in
the 65/7600's; am I vested
in having a 3 tier architecture for the foreseeable future?

On Mon, Sep 27, 2010 at 2:08 PM, Quinn Snyder  wrote:
> we are deploying them in ~50 sites (mix of 7010, 7018). smattering of
> 5k/2248 when needed. using them in a collapsed core (agg, core vdc
> model) to replace existing 650x/sup720 cores.
> running light services (eigrp, qos, multicast) but using vpc to
> provide full redundancy between 45xx/65xx closets.
> seemed like a decent choice based on lifecycle and the release of 5.0
> for the 7k.  does what we need it to do and redundancy is there.
> still feels rough, but nowhere like it used to be.
>
> q.
>
> -= sent via iphone. please excuse spelling, grammar, and brevity =-
>
> On Sep 27, 2010, at 9:32, Seth Mattinen  wrote:
>
>> About a year ago there were some large-ish threads on the Nexus and a
>> couple people that had them in production had commented that there were
>> bugs that made them feel like test subjects, plus a various assortment
>> of unexpected limitations. How much has this changed over the last year?
>>
>> I do notice that the 2248TP fabric extender supports direct to 7k, and
>> the 22xxTP datasheet lists 100/1000 as supported speeds. I've been
>> researching a 7k as a candidate for a small colo datacenter, and to me
>> it seems like it's matured quite a bit (on paper, anyway).
>>
>> ~Seth
>> ___
>> cisco-nsp mailing list  cisco-...@puck.nether.net
>> https://puck.nether.net/mailman/listinfo/cisco-nsp
>> archive at http://puck.nether.net/pipermail/cisco-nsp/
> ___
> cisco-nsp mailing list  cisco-...@puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
>

___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] How to bring one link down if another related link goes down

[John Neiberger]

On Mon, Sep 27, 2010 at 5:22 PM, Phil Mayers  wrote:
> On 09/28/2010 12:12 AM, John Neiberger wrote:
>>
>> We have two ten gig links between a pair of 7600s. For a couple of
>> reasons, if one goes down we need the other to go down automatically
>> in order to force traffic to take a completely different path instead
>> of trying to cram all of our traffic down the remaining link.
>>
>> My first thought was to use an Etherchannel with min-links set to 2.
>> However, there are architectural reasons why we may not be able to use
>> Etherchannels in this particular situation. If we can't use
>> Etherchannels, what other methods are there?
>>
>> Is there some way to utilize some sort of object tracking to force one
>> link to go down if its partner goes down? Alternatively, the link
>
> Probably the "best" option is an EEM applet. Something like:
>
> track 20 interface TenGigabitEthernet1/3 line-protocol
>  delay down 1 up 30
> event manager applet TEST
>  event track 20 state down
>  action 1.0 cli command "enable"
>  action 1.1 cli command "conf t"
>  action 1.2 cli command "int Te1/1"
>  action 1.3 cli command "shut"
>
> ...typed from memory so read the docs for more info.

That may be the best idea I've seen. It wouldn't even be necessary for
it to bring the links back up automatically. If one link goes down,
we're going to need to do a lot of careful troubleshooting before we
bring it up, so a manual restoration process would be best.

Thanks!

___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] How to bring one link down if another related link goes down

[Phil Mayers]

On 09/28/2010 12:12 AM, John Neiberger wrote:

We have two ten gig links between a pair of 7600s. For a couple of
reasons, if one goes down we need the other to go down automatically
in order to force traffic to take a completely different path instead
of trying to cram all of our traffic down the remaining link.

My first thought was to use an Etherchannel with min-links set to 2.
However, there are architectural reasons why we may not be able to use
Etherchannels in this particular situation. If we can't use
Etherchannels, what other methods are there?

Is there some way to utilize some sort of object tracking to force one
link to go down if its partner goes down? Alternatively, the link


Probably the "best" option is an EEM applet. Something like:

track 20 interface TenGigabitEthernet1/3 line-protocol
 delay down 1 up 30
event manager applet TEST
 event track 20 state down
 action 1.0 cli command "enable"
 action 1.1 cli command "conf t"
 action 1.2 cli command "int Te1/1"
 action 1.3 cli command "shut"

...typed from memory so read the docs for more info.
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

[c-nsp] How to bring one link down if another related link goes down

[John Neiberger]

We have two ten gig links between a pair of 7600s. For a couple of
reasons, if one goes down we need the other to go down automatically
in order to force traffic to take a completely different path instead
of trying to cram all of our traffic down the remaining link.

My first thought was to use an Etherchannel with min-links set to 2.
However, there are architectural reasons why we may not be able to use
Etherchannels in this particular situation. If we can't use
Etherchannels, what other methods are there?

Is there some way to utilize some sort of object tracking to force one
link to go down if its partner goes down? Alternatively, the link
itself does not need to go down. The traffic we're dealing with here
is statically routed, so it would suffice for the static routes to be
removed from the routing table unless both links were up. I have no
idea if that's possible, but we'd really like to find one or two
different ways to pull this off automatically.

Any ideas?

Thanks!
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] Cisco 4900M BGP Support

[Seth Mattinen]

On 9/27/2010 14:55, Jimmy Changa wrote:
> Thanks for the info. I'm looking for a Cisco solution that takes up the
> least amount of rack space and can handle full table. Any suggestions?
> 

I'll add another recommendation: 7201. It is basically a one-slot
7200VXR NPE-2G with 4 gig-e ports in a 1U format.

~Seth
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] Cisco 4900M BGP Support

[Chris Evans]

Asr 1000  or small 7600 would be your answer.
On Sep 27, 2010 6:04 PM, "Jimmy Changa"  wrote:
> Thanks for the info. I'm looking for a Cisco solution that takes up the
> least amount of rack space and can handle full table. Any suggestions?
>
> Jimmy Changa via Droid
> On Sep 27, 2010 5:47 PM, "Łukasz Bromirski"  wrote:
>> On 2010-09-27 23:16, Charles Klement wrote:
>>> I believe that there is only enough tcam to support about 25 ipv4
> routes.
>>
>> Exactly. The 4900M job is not to be a edge BGP router. It's a
>> TOR switch.
>>
>> As usual, one can fiddle with the incoming prefixes filtering to get
>> everything still in FIB fitting the TCAMs, it's better than 11-20k FIB
>> limit for Cat 3Ks, but it's not the ASR 1k or 7600 Sup720BXL/CXL 1mln
>> of them.
>>
>> --
>> "Everything will be okay in the end. | Łukasz Bromirski
>> If it's not okay, it's not the end." | http://lukasz.bromirski.net
>>
>>
>> ___
>> cisco-nsp mailing list cisco-nsp@puck.nether.net
>> https://puck.nether.net/mailman/listinfo/cisco-nsp
>> archive at http://puck.nether.net/pipermail/cisco-nsp/
> ___
> cisco-nsp mailing list cisco-nsp@puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] Cisco 4900M BGP Support

[Nick Hilliard]

On 27/09/2010 23:55, Jimmy Changa wrote:

Thanks for the info. I'm looking for a Cisco solution that takes up the
least amount of rack space and can handle full table. Any suggestions?


C2821 with 512k RAM.  Takes 1U vertical space.

Nick
--
(hint: it would help to specify what you're looking for in terms of 
throughput and features)

___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] Cisco 4900M BGP Support

[Łukasz Bromirski]

On 2010-09-27 23:55, Jimmy Changa wrote:

Thanks for the info. I'm looking for a Cisco solution that takes up the
least amount of rack space and can handle full table. Any suggestions?


ASR1002 with ESP10.

--
"Everything will be okay in the end.  | Łukasz Bromirski
 If it's not okay, it's not the end." |  http://lukasz.bromirski.net
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] Cisco 4900M BGP Support

[Jimmy Changa]

Thanks for the info. I'm looking for a Cisco solution that takes up the
least amount of rack space and can handle full table. Any suggestions?

Jimmy Changa via Droid
On Sep 27, 2010 5:47 PM, "Łukasz Bromirski"  wrote:
> On 2010-09-27 23:16, Charles Klement wrote:
>> I believe that there is only enough tcam to support about 25 ipv4
routes.
>
> Exactly. The 4900M job is not to be a edge BGP router. It's a
> TOR switch.
>
> As usual, one can fiddle with the incoming prefixes filtering to get
> everything still in FIB fitting the TCAMs, it's better than 11-20k FIB
> limit for Cat 3Ks, but it's not the ASR 1k or 7600 Sup720BXL/CXL 1mln
> of them.
>
> --
> "Everything will be okay in the end. | Łukasz Bromirski
> If it's not okay, it's not the end." | http://lukasz.bromirski.net
>
>
> ___
> cisco-nsp mailing list cisco-nsp@puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] Cisco 4900M BGP Support

[Łukasz Bromirski]

On 2010-09-27 23:16, Charles Klement wrote:

I believe that there is only enough tcam to support about 25 ipv4 routes.


Exactly. The 4900M job is not to be a edge BGP router. It's a
TOR switch.

As usual, one can fiddle with the incoming prefixes filtering to get
everything still in FIB fitting the TCAMs, it's better than 11-20k FIB
limit for Cat 3Ks, but it's not the ASR 1k or 7600 Sup720BXL/CXL 1mln
of them.

--
"Everything will be okay in the end.  | Łukasz Bromirski
 If it's not okay, it's not the end." |  http://lukasz.bromirski.net


___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] Cisco 4900M BGP Support

[Charles Klement]

I believe that there is only enough tcam to support about 25 ipv4 routes.

On 9/27/10, Jimmy Changa  wrote:
> Greetings,
>
>
>
> Does anyone know if the Cisco 4900M support full BGP tables? Are their
> limits to the number of routes they support? I'm running 12.2(53)SG2
> Enterprise Services, However I haven't found anything online the provides a
> definitive answer.
>
> ___
> cisco-nsp mailing list  cisco-nsp@puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
>
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

[c-nsp] Cisco 4900M BGP Support

[Jimmy Changa]

Greetings,

 

Does anyone know if the Cisco 4900M support full BGP tables? Are their
limits to the number of routes they support? I'm running 12.2(53)SG2
Enterprise Services, However I haven't found anything online the provides a
definitive answer.

___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] Nexus evolution

[Quinn Snyder]

we are deploying them in ~50 sites (mix of 7010, 7018). smattering of
5k/2248 when needed. using them in a collapsed core (agg, core vdc
model) to replace existing 650x/sup720 cores.
running light services (eigrp, qos, multicast) but using vpc to
provide full redundancy between 45xx/65xx closets.
seemed like a decent choice based on lifecycle and the release of 5.0
for the 7k.  does what we need it to do and redundancy is there.
still feels rough, but nowhere like it used to be.

q.

-= sent via iphone. please excuse spelling, grammar, and brevity =-

On Sep 27, 2010, at 9:32, Seth Mattinen  wrote:

> About a year ago there were some large-ish threads on the Nexus and a
> couple people that had them in production had commented that there were
> bugs that made them feel like test subjects, plus a various assortment
> of unexpected limitations. How much has this changed over the last year?
>
> I do notice that the 2248TP fabric extender supports direct to 7k, and
> the 22xxTP datasheet lists 100/1000 as supported speeds. I've been
> researching a 7k as a candidate for a small colo datacenter, and to me
> it seems like it's matured quite a bit (on paper, anyway).
>
> ~Seth
> ___
> cisco-nsp mailing list  cisco-nsp@puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] Nexus evolution

[Tim Stevenson]

Next major s/w release (Cairo, release # most likely to be 5.1) 
supports 2248 to n7k directly. 2232 comes a bit later (within 6-8 months).


Hope that helps,
Tim


At 09:45 AM 9/27/2010, David Freedman declared:

I believe that this direct-to-7k support is only just being released in
s/w (aug/sep) and it will be limted to 32 FEX per 7k (and fex must be
2248 or 2232, 2148 not supported)





Tim Stevenson, tstev...@cisco.com
Routing & Switching CCIE #5561
Distinguished Technical Marketing Engineer, Cisco Nexus 7000
Cisco - http://www.cisco.com
IP Phone: 408-526-6759

The contents of this message may be *Cisco Confidential*
and are intended for the specified recipients only.


___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] Nexus evolution

[David Freedman]

Seth Mattinen wrote:
> About a year ago there were some large-ish threads on the Nexus and a
> couple people that had them in production had commented that there were
> bugs that made them feel like test subjects, plus a various assortment
> of unexpected limitations. How much has this changed over the last year?
> 
> I do notice that the 2248TP fabric extender supports direct to 7k, and
> the 22xxTP datasheet lists 100/1000 as supported speeds. I've been
> researching a 7k as a candidate for a small colo datacenter, and to me
> it seems like it's matured quite a bit (on paper, anyway).
> 

I believe that this direct-to-7k support is only just being released in
s/w (aug/sep) and it will be limted to 32 FEX per 7k (and fex must be
2248 or 2232, 2148 not supported)

If this doesn't work for you then you need to retain your 5k agg layer.

Dave.

> ~Seth
> ___
> cisco-nsp mailing list  cisco-nsp@puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
> 


-- 


David Freedman
Group Network Engineering
Claranet Group

___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] ISR G2 performance

[Peter Rathlev]

On Mon, 2010-09-27 at 11:33 -0400, Keegan Holley wrote:
> I'm a little annoyed by their stance though.  I just want them to make a
> recommendation that I can use instead of trying to fill my head with
> marketing nonsense.

I personally don't see "routerperformance.pdf" as marketing nonsense. It
gives you a basic figure to work with and makes it possible to compare
different platforms.

Of course a set of standard cases could be documented, e.g. "simple NAT,
one inside and one outside interface" or "simple LLQ with this specific
configuration". But one man's standard setup is an exotic setup for many
of his colleagues.

If Cisco were to announce "best case" forwarding figures, I would call
that marketing nonsense. :-)

-- 
Peter


___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] Nexus evolution

[Chris Evans]

We are deploying 7-5-2 like mad.

Stable platform for its age.
On Sep 27, 2010 12:34 PM, "Seth Mattinen"  wrote:
> About a year ago there were some large-ish threads on the Nexus and a
> couple people that had them in production had commented that there were
> bugs that made them feel like test subjects, plus a various assortment
> of unexpected limitations. How much has this changed over the last year?
>
> I do notice that the 2248TP fabric extender supports direct to 7k, and
> the 22xxTP datasheet lists 100/1000 as supported speeds. I've been
> researching a 7k as a candidate for a small colo datacenter, and to me
> it seems like it's matured quite a bit (on paper, anyway).
>
> ~Seth
> ___
> cisco-nsp mailing list cisco-nsp@puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] cisco MPLS AutoBandwidth Allocator

[Peter Rathlev]

On Mon, 2010-09-27 at 09:21 +0530, jack daniels wrote:
> I'm stuck in the understanding - does oit make sense to implement MPLS
> AutoBandwidth in scenario where I have only 20 subnets max to be sent
> on this Backbone.

AutoBandwidth gives you the advantage of having the network recalculate
LSPs every now and then. This can (partly) overcome the scenario where
the specific time of establishing an LSP would have a negative effect on
how it's built.

If you use TE extensively and have one or more paths that are
oversubscribed (via RSVP) then AutoBandwidth may be used to shift around
some of the paths to (maybe) achieve a better utilisation.

If you don't have overlapping TE tunnels there's no point in using
AutoBandwidth. If you have plenty capacity (i.e. the sum of bandwidth of
all tunnels is less than you "narrowest" link's capacity) there's also
no need IMO.

I your network is carefully engineered off-line and you never want the
network to change you engineering you also don't need AutoBandwidth.

-- 
Peter


___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

[c-nsp] Nexus evolution

[Seth Mattinen]

About a year ago there were some large-ish threads on the Nexus and a
couple people that had them in production had commented that there were
bugs that made them feel like test subjects, plus a various assortment
of unexpected limitations. How much has this changed over the last year?

I do notice that the 2248TP fabric extender supports direct to 7k, and
the 22xxTP datasheet lists 100/1000 as supported speeds. I've been
researching a 7k as a candidate for a small colo datacenter, and to me
it seems like it's matured quite a bit (on paper, anyway).

~Seth
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] SFP info

[Ingen Schenau, Jeroen van (ICTS)]

> If I use SFP LX connector and coonect two routers ports back to back.
> Will this have any impact on router port .

No impact, it should work. You can connect LX transceivers back to back;
iirc, minimum patch length is somewhere around 0.5 to 2 meters.

> Is that I can use only SX SFP for Back to back connectivity with routers.

No, LX or BX are OK too. Only with ZX or other long-haul optics you
might need an attenuator.

> Also if my router port is coonecting to MUX which is colocated , I
> need to mandatorly use SX SFP.

That depends on what the MUX supports, I guess... SX is generally used
on short distances since the optics are cheaper.


Regards,

Jeroen van Ingen
ICT Service Centre
University of Twente, P.O.Box 217, 7500 AE Enschede, The

___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] ISR G2 performance

[Keegan Holley]

I'm a little annoyed by their stance though.  I just want them to make a
recommendation that I can use instead of trying to fill my head with
marketing nonsense.  They certainly will not refund my money if I make the
wrong decision.  Even with the 2800 series, they made the recommendation but
buried it on the Q&A page.  Those numbers weren't even worse case, those
were the recommended uses.  Thanks for the docs though.

Keegan


On Mon, Sep 27, 2010 at 5:35 AM, Stephen Stack
wrote:

> Heyho,
>
> Ran into some issues around this not too long ago.
>
> I spoke to my local SE at the time about it, and his line
> Was basically - what the marketing bumff says - stands!!!
> i.e. in my case, a Cisco 2951 was rated for 75Mbps
>
> If you turn on all the services - this may be the case - but who does
> that???
> (recently listened to a podcast where Cisco where saying that there are
> over
> 4000 separate services in IOS)
>
> I gave him my requirements of Ethernet-Ethernet connectivity with some
> static routing and with some HQOS (shaping)
> The stats he gave me back from their internal testing was much higher
> that 75 Mbps.
>
> Cisco, being the engineering company they are, need to divulge much
> better
> Information and more real world stats than marketing figures to help
> Out us lowly engineers in the field when making decisions on kit.
>
> I found some information for you to make some informed decisions.
> Try this URL for some Math on ISR G1
> https://supportforums.cisco.com/message/714842#714842
>
> And this for Cisco Internal figures on ISR G2s
>
> "Ask your Cisco partner to give you access to the following document:
> http://www.cisco.com/en/US/partner/prod/collateral/routers/ps10536/white
> _paper_c11_595485.pdf
> Since that is proprietary information, I'm not allowed to post it here."
>
>
> -Original Message-
> From: cisco-nsp-boun...@puck.nether.net
> [mailto:cisco-nsp-boun...@puck.nether.net] On Behalf Of Keegan Holley
> Sent: 27 September 2010 06:07
> To: Cisco NSPs
> Subject: [c-nsp] ISR G2 performance
>
> Do these numbers seem a little off to anyone else?  I'm shopping again
> and
> I'm trying to compare the ISR G2's with the mainstay 3845 for a DS3
> site.
>  As usual it has been hard to find performance numbers on cisco.com.
> Soon
> I'm expecting routers that walk into the DC and install themselves only
> to
> perform at 3Mbps, but I digress...  A little googling brought me the
> following link, but the numbers seem a little suspect.  For example the
> entire 2800 series is rated above 45M in the document, but the product
> Q&A
> recommends the 2851 just to aggregate 6 T1's.  The pages for the modules
> are
> no less misleading.  Is there a way to find performance metrics for
> these
> routers stated in plain terms?
>
> http://www.cisco.com/web/partners/downloads/765/tools/quickreference/rou
> terperformance.pdf
>
> http://www.cisco.com/en/US/prod/collateral/routers/ps5854/prod_qas0900ae
> cd80169bd6.html
> ___
> cisco-nsp mailing list  cisco-nsp@puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
>
>
>
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

[c-nsp] ISG DHCP and PPPoE in the same vlan

[MKS]

Hi list

I was wondering if it's possible to do both PPPoE and (ISG) DHCP
subscriber termination in the same vlan on e.g. a 7200 ?

What about L3 addresses in the DHCP model, can I somehow use the same
L3 address/pool for 100 vlans without opening up L2 user-to-user
communication.

Regards
MKS
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

[c-nsp] SFP info

[jack daniels]

Hi guys,


If I use SFP LX connector and coonect two routers ports back to back.
Will this have any impact on router port .

Is that I can use only SX SFP for Back to back connectivity with routers.


Also if my router port is coonecting to MUX which is colocated , I
need to mandatorly use SX SFP.


Regards
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

[c-nsp] SFP info

[jack daniels]

Hi guys,


If I use SFP LX connector and coonect two routers ports back to back.
Will this have any impact on router port .

Is that I can use only SX SFP for Back to back connectivity with routers.


Also if my router port is coonecting to MUX which is colocated , I
need to mandatorly use SX SFP.


Regards
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] vlan problem in my router

[Justin M. Streiner]

On Mon, 27 Sep 2010, Deric Kwok wrote:


I have problem to create vlan in my router 28 serise.

I can create vlan database

1/ I can't use name command as "router(vlan)# name net1"

2/ I can't config t the vlan as "(config)#vlan 10  ===> % Invalid
input detected at '^' marker."

ls my router supporting vlan?


If you are running an IOS version that supports 802.1Q VLAN tagging, you 
can create VLANs, but they behave differently on a router than on a 
switch:


For example, on a router:

interface GigabitEthernet0/0/0
 no ip address
interface GigabitEthernet0/0/0.100
 description sub-interface for VLAN 100
 encapsulation dot1q 100
 ip address 10.10.100.1 255.255.255.0
interface GigabitEthernet0/0/0.200
 description sub-interface for VLAN 200
 encapsulation dot1q 200
 ip address 10.10.200.1 255.255.255.0

You can shut down individual sub-interfaces if you need to, and as you'd 
expect, shutting down the parent interface (Gig0/0/0 in this case) will 
take down all of the sub-interfaces.


You need to pay attention to which VLAN you use as the native VLAN on 
your trunk, and match it on the router by adding the "native" statement to 
your subinterface encapsulation command.  There are other things that need 
to match up as well, such as the encapsulation type, especially if you're 
still using something other than 802.1q.  I'm not sure if ISL is even 
supported anymore.


There are other important things to consider as well, such as which device 
acts as the root bridge for the VLANs on the trunk.


There are lots of examples and documentation for configuring VLANs and 
trunks on routers on Cisco's website.


jms
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] DS3 Nubie

[Jeff Wojciechowski]

Hi Jon-

Kind of what I was thinking - if the transport (fiber or other transport) is on 
the other side of a managed Ethernet box I won't know if there are physical 
line problems as I won't be able to see interface counters on it...

As it is right now I get emailed anytime any interface on my network takes 
errors I would be leery of any connection that I couldn't see what errors were 
incrementing on the link between CO to my site.

Thanks,

-Jeff


-Original Message-
From: cisco-nsp-boun...@puck.nether.net 
[mailto:cisco-nsp-boun...@puck.nether.net] On Behalf Of Jon Lewis
Sent: Sunday, September 26, 2010 8:32 AM
To: Christopher J. Wargaski
Cc: cisco-nsp@puck.nether.net
Subject: Re: [c-nsp] DS3 Nubie

On Sat, 25 Sep 2010, Christopher J. Wargaski wrote:

>   This year I installed a video WAN comprised of several 3845 routers
> with the NM-1T3/E3 for point to point DS3s (that is all, nothing
> else). The 3845 routers list at $13,000 and the network module at
> $8,500. You should do yourself and your firm a favor and look at
> business class Ethernet. DS3s are so expensive and sometimes a major
> pain in the butt.
>
>   For business class Ethernet, you most major carriers can offer you
> something. Typically, fiber is pulled to your NetPOP and a switch is
> installed, your hand-off is a switch port. The downside is the cost of
> delivering fiber to your building which can often be quite prohibitive
> unless you can amortize that cost over a number of years.
> (Unfortunately, that may not be possible since the deliverable may not
> be considered a product to your accounting department.)

There are other downsides [to metro ethernet] to consider.  Using ethernet for 
long haul, your devices at each end are no longer directly connected, but will 
have a network (the provider's ethernet switches) between them.
Failures in "the network" will cause your ends to lose contact with each other, 
while their interfaces remain up/up.

Depending on the speed provisioned by the provider, and the speed of your 
networks, you can run into bursting/packet drops issues when your 1000baseT 
traffic hits the 10mbit or 100mbit metro-E.

Also, because your packets are flowing over the provider's switched network, 
they typically ride a specific VLAN.  If (no, when) the provider screws up and 
puts another customer in the same VLAN, very strange things will happen, 
particularly if you're both using the same RFC1918 IPs.
This is something you generally don't have to worry about with private lines.

--
  Jon Lewis, MCP :)   |  I route
  Senior Network Engineer |  therefore you are
  Atlantic Net|
_ http://www.lewis.org/~jlewis/pgp for PGP public key_ 
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net 
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

This electronic mail (including any attachments) may contain information that 
is privileged, confidential, or otherwise protected from disclosure to anyone 
other than its intended recipient(s). Any dissemination or use of this 
electronic mail or its contents (including any attachments) by persons other 
than the intended recipient(s) is strictly prohibited. If you have received 
this message in error, please delete the original message in its entirety 
(including any attachments) and notify us immediately by reply email so that we 
may correct our internal records.  Midland Paper Company accepts no 
responsibility for any loss or damage from use of this electronic mail, 
including any damage resulting from a computer virus.

___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

[c-nsp] Bondig in 878

[Emilio Balmon]

Hi all,

How can i know what is the real speed that synchronizing my bundle line 
with two shdsl 2M/2M?. It's working fin?  Add the output of sh 
controller dsl 0:


CPE251#sh controllers dsl 0

DSL 0 controller UP
Globespan xDSL controller chipset
Line Mode: Four Wire Standard
DSL mode: Trained with SHDSL Annex B
Frame mode: Utopia
Configured Line rate: 4096Kbps
Line Re-activated 1 times after system bootup
LOSW Defect alarm: ACTIVE
CRC per second alarm: ACTIVE
Line termination: CPE

Line 0 statistics

Current 15 min counters
CRC : 0 LOSW Defect : 0 ES : 0  SES : 0 UAS : 0

Previous 15 min counters
CRC : 0 LOSW Defect : 0 ES : 0  SES : 0 UAS : 0

Current 24 hr counters
CRC : 5 LOSW Defect : 0 ES : 5  SES : 0 UAS : 32

Previous 24 hr counters
CRC : 5 LOSW Defect : 0 ES : 5  SES : 0 UAS : 0


Line 1 statistics

Current 15 min counters
CRC : 0 LOSW Defect : 0 ES : 0  SES : 0 UAS : 0

Previous 15 min counters
CRC : 0 LOSW Defect : 0 ES : 0  SES : 0 UAS : 0

Current 24 hr counters
CRC : 0 LOSW Defect : 0 ES : 0  SES : 0 UAS : 32

Previous 24 hr counters
CRC : 0 LOSW Defect : 0 ES : 0  SES : 0 UAS : 0

Line-0 status
Chipset Version:  0
Firmware Version:  R4.2.1
Modem Status:  Data, Status 1
Last Fail Mode:  Failed during Carrier Detect status:0x1
Line rate:  2056 Kbps
Framer Sync Status: In Sync
Rcv Clock Status: In the Range
Loop Attenuation:  13.5 dB
Transmit Power:  14.5 dB
Receiver Gain:  27.7420 dB
SNR Sampling:  37.9860 dB
Receive HEC Error Count:  0
Line-1 status
Chipset Version:  0
Firmware Version:  R4.2.1
Modem Status:  Data, Status 1
Last Fail Mode:  Failed during Carrier Detect status:0x1
Line rate:  2056 Kbps
Framer Sync Status: In Sync
Rcv Clock Status: In the Range
Loop Attenuation:  13.1 dB
Transmit Power:  14.5 dB
Receiver Gain:  27.420 dB
SNR Sampling:  38.9550 dB
Receive HEC Error Count:  0
Dying Gasp: Present

Thanks in advanced


--




**Fax. +34 91 360 14 71
Paseo de la Castellana 21-28046 Madrid (Spain) *www.gowex.com 
* 	

Síguenos en:

*AVISO/WARNING*
"Este mensaje y los documentos que, en su caso, lleve anexos, pueden 
contener información confidencial. Por ello, se informa a quien lo 
reciba por error que la información contenida en el mismo es reservada y 
su uso no autorizado está prohibido legalmente, por lo que en tal caso 
le rogamos que nos lo comunique por la misma vía o por teléfono (+34 91 
360 14 70), se abstenga de realizar copias del mensaje o remitirlo o 
entregarlo a otra persona y proceda a barlo de inmediato."
Confidential Information may be contained in this message and any of its 
annexes If you are not the addressee indicated in this message you are 
here by informed that such information is private and any unauthorized 
use of said information is forbidden. You may not copy or deliver this 
message to anyone, you should destroy it immediately and kindly notify 
the sender by reply email or by telephone (+34 91 360 14 70)".



___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

[c-nsp] vlan problem in my router

[Deric Kwok]

Hi

I have problem to create vlan in my router 28 serise.

I can create vlan database

1/ I can't use name command as "router(vlan)# name net1"

2/ I can't config t the vlan as "(config)#vlan 10  ===> % Invalid
input detected at '^' marker."

ls my router supporting vlan?

if not, how I can create vlan as my router is only 2 ports.

Thank you
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] ISR G2 performance

[Stephen Stack]

Heyho,

Ran into some issues around this not too long ago. 

I spoke to my local SE at the time about it, and his line
Was basically - what the marketing bumff says - stands!!!
i.e. in my case, a Cisco 2951 was rated for 75Mbps 

If you turn on all the services - this may be the case - but who does
that???
(recently listened to a podcast where Cisco where saying that there are
over
4000 separate services in IOS)

I gave him my requirements of Ethernet-Ethernet connectivity with some
static routing and with some HQOS (shaping)
The stats he gave me back from their internal testing was much higher
that 75 Mbps.

Cisco, being the engineering company they are, need to divulge much
better
Information and more real world stats than marketing figures to help
Out us lowly engineers in the field when making decisions on kit.

I found some information for you to make some informed decisions. 
Try this URL for some Math on ISR G1
https://supportforums.cisco.com/message/714842#714842

And this for Cisco Internal figures on ISR G2s

"Ask your Cisco partner to give you access to the following document:
http://www.cisco.com/en/US/partner/prod/collateral/routers/ps10536/white
_paper_c11_595485.pdf
Since that is proprietary information, I'm not allowed to post it here."


-Original Message-
From: cisco-nsp-boun...@puck.nether.net
[mailto:cisco-nsp-boun...@puck.nether.net] On Behalf Of Keegan Holley
Sent: 27 September 2010 06:07
To: Cisco NSPs
Subject: [c-nsp] ISR G2 performance

Do these numbers seem a little off to anyone else?  I'm shopping again
and
I'm trying to compare the ISR G2's with the mainstay 3845 for a DS3
site.
 As usual it has been hard to find performance numbers on cisco.com.
Soon
I'm expecting routers that walk into the DC and install themselves only
to
perform at 3Mbps, but I digress...  A little googling brought me the
following link, but the numbers seem a little suspect.  For example the
entire 2800 series is rated above 45M in the document, but the product
Q&A
recommends the 2851 just to aggregate 6 T1's.  The pages for the modules
are
no less misleading.  Is there a way to find performance metrics for
these
routers stated in plain terms?

http://www.cisco.com/web/partners/downloads/765/tools/quickreference/rou
terperformance.pdf

http://www.cisco.com/en/US/prod/collateral/routers/ps5854/prod_qas0900ae
cd80169bd6.html
___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

Re: [c-nsp] cisco MPLS AutoBandwidth Allocator

[jack daniels]

its core transport network for connectivity betweeen GSM packet core devices.



On Mon, Sep 27, 2010 at 10:29 AM, Keegan Holley
 wrote:
> Unless all 20 subnets are exactly the same, you may not need mpls at all.
>  Why did you go with MPLS in the first place?  What sort of network is it?
>
>
> On Sun, Sep 26, 2010 at 11:51 PM, jack daniels 
> wrote:
>>
>> Hi guys,
>>
>> I'm stuck in the understanding - does oit make sense to implement MPLS
>> AutoBandwidth in scenario where I have only 20 subnets max to be sent
>> on this Backbone. What will be pros and corns of this , if u experts
>> can show me a path.
>>
>>
>> Thanks and Regards
>>
>>
>> On Fri, Aug 27, 2010 at 1:26 AM, Peter Rathlev  wrote:
>> > On Fri, 2010-08-27 at 00:21 +0530, jack daniels wrote:
>> >> hope MPLS TE - Auto-bandwidth enhancement  and MPLS TE -
>> >> Auto-bandwidth allocator are same feature
>> >
>> > That seems strange. So you're saying you don't need something technical
>> > solution to a problem you have, you need a feature called
>> > "Auto-bandwidth allocator"? Even if Cisco defined that to be something
>> > controlling the LEDs on your front panels and nothing else? :-)
>> >
>> > Googling "Auto-bandwidth allocator" gives a link to a document
>> > (apparantly from 2006) describing the feature for 7500 and 7200.
>> > Skimming that document and the more recent one about the 7600 points
>> > towards the latter being an evolved version of the former. That took me
>> > about 2 minutes at most. I shouldn't have.
>> >
>> > --
>> > Peter
>> >
>> >
>> >
>>
>> ___
>> cisco-nsp mailing list  cisco-...@puck.nether.net
>> https://puck.nether.net/mailman/listinfo/cisco-nsp
>> archive at http://puck.nether.net/pipermail/cisco-nsp/
>>
>>
>
>

___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/