[Outages-discussion] [outages] Equinix Exchange contacts

[James Cornman]

Thanks for moving. I was unaware of this list.

In my opinion, this requirement is a bit extraneous, however the downstream
customer with this mandate is pretty passionate about this configuration.
That being said, the benefit is that if any peer has a preference for
either one of our next hops (we cant control their route selection to
ensure they do ECMP), we still get the upside of the extra capacity, as
well as port level redundancy. I can't imagine having 1 more mac address in
the exchange point's port-security criteria is overly complicated. That
apparently is the biggest point of contention with Equinix thus far, since
its "non standard". To note, we're not even requiring different chassis on
the exchange point side; just want to take advantage of the MC LAG on our
side so from the Equinix side, its pretty simple to make us go away :)

-James



On Tue, Nov 21, 2017 at 11:38 AM, Hugo Slabbert <hslabbert at demonware.net>
wrote:

> Moved to outages-discussion.
>
> This seems like outages-discussion at outages.org at best and likely more
> suited to nanog.  I don't see anything related to an outage on this.
>
> On the surface, why not do discrete L3 links and equal-cost across the
> members rather than shoving L2 redundancy and complexity (MC-LAG) in the
> middle only to stuff L3 redundancy back over top of it (multiple IPs &
> discrete routing protocol sessions)?
>
> --
> Hugo
>
> On Tue 2017-Nov-21 11:14:51 -0500, Chad Austin via Outages <
> outages at outages.org> wrote:
>
> Can you describe what the end result would look like?
>>
>> -Chad
>>
>>
>> On 11/21/2017 11:11 AM, James Cornman via Outages wrote:
>>
>>> Anyone present with some contacts / influence with Equinix regarding
>>> their Exchange product?
>>>
>>> We're working to establish what they consider a unique situation thats
>>> not supported, and we're mostly getting told NO from the Equinix NOC w/o
>>> any escalation, and I'm looking to get someone with more technical
>>> expertise to talk to.
>>>
>>> We're looking to establish 2x10Gbps connections in an LACP LAG, with 2
>>> IP addresses; 1 per MC LAG member, so we get the benefit of redundancy and
>>> capacity. Apparently the 2 IP address thing is the current show stopper. I
>>> haven't received any meaningful feedback as to whether this is a political
>>> problem or a technical one.
>>>
>>> FWIW, Other exchange points support this such as Any2 and NYIIX, etc.
>>>
>>> -James
>>>
>>> --
>>>
>>> *James Cornman*
>>>
>>> /Chief Technology Officer
>>> /jcornman at atlanticmetro.net <mailto:jcornman at atlanticmetro.net>
>>> 212.792.9950 - ext 101
>>> *
>>> Atlantic Metro Communications*
>>>
>>> /4 Century Drive, Parsippany NJ  07054/
>>>
>>> /
>>> //Cloud Hosting //• //Colocation • Network Connectivity • Managed
>>> Services/
>>>
>>> Follow us on Twitter: @atlanticmetro <https://twitter.com/atlanticmetro>
>>> /• Like us on Facebook <https://www.facebook.com/atlanticmetro>/
>>> www.atlanticmetro.net <https://www.atlanticmetro.net/>
>>>
>>>
>>>
>>> _______________________________________________
>>> Outages mailing list
>>> Outages at outages.org
>>> https://puck.nether.net/mailman/listinfo/outages
>>>
>>
>>
> _______________________________________________
>> Outages mailing list
>> Outages at outages.org
>> https://puck.nether.net/mailman/listinfo/outages
>>
>
>


-- 

*James Cornman*


*Chief Technology Officer*jcornman at atlanticmetro.net
212.792.9950 - ext 101

*Atlantic Metro Communications*

*4 Century Drive, Parsippany NJ  07054*


*Cloud Hosting **• **Colocation • Network Connectivity • Managed Services*

Follow us on Twitter: @atlanticmetro <https://twitter.com/atlanticmetro> *•
Like us on Facebook <https://www.facebook.com/atlanticmetro>*
www.atlanticmetro.net
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://puck.nether.net/pipermail/outages-discussion/attachments/20171121/b4820fda/attachment.html>