[j-nsp] What do you think about the MX line?

[Keegan Holley]

2011/6/26 Mark Tinka <mtinka at globaltransit.net>

> On Monday, June 27, 2011 06:56:48 AM Keegan Holley wrote:
>
> > I think the general attitude is positive towards them.
> > They are a good compliment to the M/T series and
> > generally solid flexible boxes.  You should probably
> > include how you plan to use them in your question.  For
> > example a few list members complain about multicast/IGMP
> > bugs and other issues with the new trio based cards and
> > some of the new code.  If you don't run alot of
> > multicast these wouldn't really apply to you.
>
> For us, we use them heavily in the edge, and that hasn't
> been the smoothest of rides.
>
> My guess is if you need them for peering or in the core, you
> might have less issues, but not necessarily (we already know
> of core applications where the MX could be troublesome - but
> the edge role takes the cake, by far).
>
>
Please elaborate.  This is interesting since most edge boxes have the fewest
features configured.  Is this an metro-e edge or just vanilla IP?  V4/v6?


> There are also some limitations, so far, if we use them as
> BRAS's, but these are mostly bugs are feature unavailability
>

Other than the obvious I'm at a loss for the definition of "BRAS".


> at this time. The problem is that without the feature being
> present today, it's hard to know how the box will scale,
> which could be a big problem unto itself.
>
> All in all, it depends on the complexity/sophistication of
> your deployment, the role you're placing the MX in, and what
> features you're going to need. For some folk, it's the
> perfect box. For others, it's less so.
>
> Mark.
>
>