[c-nsp] data center/mpls/vpls

Aaron aaron1 at gvtc.com
Tue Apr 16 10:44:07 EDT 2013


Thanks for the warning on the 9000v

We have thought about using it but aren't going to for the DC deployment.
(it doesn't have the (11) 10gig interfaces we need to begin with)  It always
seemed attractive that it was advertised as a linecard in an asr9k with all
the features that an asr9k/ios xr has to offer, but your commentary doesn't
make me feel very good about it

Jared, why do you say it's IOS ?  If it's a linecard in an ASR9k wouldn't
that make it IOS XR?  Not sure what you mean by ios

Also, off the top of your heard, are there problems with IOS XR 4.3.1 in
ASR9k that I should be aware of ?

Aaron

p.s. hope y'all don't mind, I'm adding the list back to cc's, since this
seems like good info for the community to benefit from.

-----Original Message-----
From: Jared Mauch [mailto:jared at puck.nether.net] 
Sent: Tuesday, April 16, 2013 6:41 AM
To: Gabor Szabo (gabszabo)
Cc: Aaron; Oliver Garraux; Jeff Kell
Subject: Re: [c-nsp] data center/mpls/vpls

You might want to be careful with the 9000V.  There are a number of
interesting side-effects of it.

It basically won't work right until you get to at least 4.3.1.  They are
missing a lot of things.  It is also an IOS box, which means you get none of
the benefits of IOS-XR.  They hacked it together using TFTP and a VLAN tag.
The inventory doesn't work right after 3 releases (4.2.2, 4.2.3, 4.3.0) and
the software management on it is not under the ADMIN plane.

Basically, another cisco half-solution, so use with caution.

- Jared


From: Mick O'Rourke [mailto:mkorourke at gmail.com] 
Sent: Tuesday, April 16, 2013 6:36 AM
To: Gabor Szabo (gabszabo)
Cc: Aaron; Oliver Garraux; Jeff Kell; cisco-nsp at puck.nether.net
Subject: Re: [c-nsp] data center/mpls/vpls

Wouldn't go near a 9000v it for a data centre deployment on AC power  - no
dual AC option.


On Apr 16, 2013, at 7:14 AM, "Gabor Szabo (gabszabo)" <gabszabo at cisco.com>
wrote:

> Hi Aaron,
> 
> Have you checked the ASR9K nV Satellite feature with 9000v which can be
placed physically as TOR device but works as a remote linecard for the
ASR9K? You can have every service on satellite ports what you have on local
interface of the ASR9K...
> 
> It is generally not recommended / preferred architecture for hosting /
complex DC (where the Nexus family is our strong preference) but can fit for
collocation type of services...
> 
> Regards,
> Gabor
> 
>> -----Original Message-----
>> From: cisco-nsp [mailto:cisco-nsp-bounces at puck.nether.net] On Behalf 
>> Of Aaron
>> Sent: Monday, April 15, 2013 10:27 PM
>> To: 'Oliver Garraux'; 'Jeff Kell'
>> Cc: cisco-nsp at puck.nether.net
>> Subject: Re: [c-nsp] data center/mpls/vpls
>> 
>> Thanks Oliver for the input, I'm planning on not doing L3 on 
>> whichever DC ToR switch I go with.... and simply doing the L3 on the 
>> uplinked ASR9006
>> 
>> Aaron
>> 
>> -----Original Message-----
>> From: cisco-nsp [mailto:cisco-nsp-bounces at puck.nether.net] On Behalf 
>> Of Oliver Garraux
>> Sent: Monday, April 15, 2013 3:51 PM
>> To: Jeff Kell
>> Cc: cisco-nsp at puck.nether.net
>> Subject: Re: [c-nsp] data center/mpls/vpls
>> 
>> At a large enterprise, we're using them for L2 only.  For a smallish 
>> virtualized hosting environment we're doing L3 with them.  I think 
>> the people working on the hosting environment are happy with them.  
>> My concerns about L3 on the N5k are mostly about some of the limitations.
Off the top of my head:
>> 
>> - you can't do ISSU with L3
>> - it doesn't support PBR
>> - since a port-channel is used internally to connect to the L3 
>> module, you may not be able to effectively use 160 gbps for L3
>> - more limited # of FEX's supported when L3 is being used.  I think 
>> they increased it though in some version of code, so this might be 
>> less of an issue today
>> 
>> If you can live with the limitations, I think L3 on the N5K is probably
OK.
>> Not sure that its the best choice for L3 stuff though.
>> 
>> Oliver
>> 
>> -------------------------------------
>> 
>> Oliver Garraux
>> Check out my blog:  blog.garraux.net
>> Follow me on Twitter:  twitter.com/olivergarraux
>> 
>> 
>> On Mon, Apr 15, 2013 at 2:56 PM, Jeff Kell <jeff-kell at utc.edu> wrote:
>> 
>>> On 4/15/2013 2:07 AM, Andrew Miehs wrote:
>>>>> I would like a pair of top-of-rack devices that can mirror one 
>>>>> another
>>> if
>>>>> possible too, like cisco's 6509-vss thing.or at least like nexus 
>>>>> vpc (multichassis link aggregation/bundling)
>>>>> 
>>>> The Nexus 5Ks are pretty cheap and good if you only need L2.
>>> 
>>> I'm still a Nexus virgin... so excuse my naive question... but... :)
>>> 
>>> I've heard more than once that Nexus 5K is not quite ready for 
>>> Layer-3 prime time, but I have also heard others swear by at (as 
>>> opposed to swearing AT it).
>>> 
>>> So what's the real scoop?  Are you deploying N5K just for L2 and 
>>> front-ending it with some other L3 device?
>>> 
>>> Jeff
>>> 
>>> _______________________________________________
>>> 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/
>>> 
>> _______________________________________________
>> 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/
>> 
>> _______________________________________________
>> 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/
> 
> _______________________________________________
> 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/



More information about the cisco-nsp mailing list