[j-nsp] J Series - BGP Peering Router?

[Chen Jiang]

Per Juniper's tradition, datasheet numbers just means tested by QA but not
the system limitation.

JUNOS in control panel doesn't pre-define any limitation for routing
capacity, it just depneding on the memory capacity.

- Show quoted text -
On Fri, Apr 23, 2010 at 5:33 AM, Richard A Steenbergen <ras at e-gerbil.net>
wrote:

On Thu, Apr 22, 2010 at 04:07:35PM -0400, Paul Stewart wrote:
> Thanks very much for the feedback.. I've received a few offline replies as
> well.
>
> To clarify the "small nature" of this application - we would be sending
> about 65 iBGP routes to the box and receiving about 4000 eBGP routes.
It's
> for a small peering pop/exchange point.


Ah ok, when I see 65 peers I'm used to it being several million paths.
I'm sure 4k or even 20k routes are perfectly fine, just watch out for
the performance issues if you want to start stuffing many full tables
into a J-series.


--
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)
_______________________________________________
juniper-nsp mailing list juniper-nsp at puck.nether.net
https://puck.nether.net/mailman/listinfo/juniper-nsp





-- 
BR!

On Fri, Apr 23, 2010 at 12:30 AM, Paul Stewart <paul at paulstewart.org> wrote:

> Hi there..
>
>
>
> I have a couple of applications pop up recently where I think a J-Series
> might suffice for BGP peering.  The application is a small peering POP
> doing
> about 200Mb/s of traffic, about 50 BGP peers, and total routes is roughly
> 4000 total.
>
>
>
> Would a J2320 with 1 GB DRAM (which I just happen to have a spare with
> J-Care already on it) suffice?  I'm reading the datasheet on J-Series
> currently and it shows me limitations such as:
>
>
>
> BGP instances: 32 (is this separate ASN's?)
>
> BGP peers: 64 (is that hard coded or only a recommendation?)
>
> BGP routes (400k - perfect, what more can I say)
>
>
>
> Any input is much appreciated .. especially performance numbers and limits
> around peers themselves.  We are slowly deploying Juniper but have no BGP
> speaking boxes up yet....
>
>
>
> Thanks,
>
>
>
> Paul
>
>
>
> _______________________________________________
> juniper-nsp mailing list juniper-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/juniper-nsp
>



-- 
BR!



          James Chen