Re: [armd] Multicast in the data center [was Re: address resolution

Hi Linda,

Yes, pub-sub is the same as publish-subscribe.  The publishers and subscribers are both in the same subnet and across subnets, even on opposite sides of of a WAN.  In fact many of the applications both publish and subscribe simultaneously.  Some applications use it for heart-beat and as bi-directional application control channels, while others use it to send fat (50Mbps to 1Gbps+) one-way data flows to many downstream recipients, and everything in between.  Multicast is generally organized into small, medium and large flow groups to manage state.  Small flows are generally ad-hoc (come and go at will) and forced to share a limited set of small flow groups.  Medium flow groups are shared by communities of interest.  Large flows get their own multicast groups to avoid excessive bandwidth use.  A PGM layer is used for reliable multicast.  Others will have different use-cases and operating models.  Bottom line -- multicast IS being used extensively in the DC, but not by run-of-the-mill applications.

-- aldrin

On Feb 15, 2012, at 9:53 AM, Linda Dunbar wrote:

> Aldrin, 
> 
> Is "pub-sub" same as "Publish-subscribe" related applications? 
> Are all the "subscribers" in the same subnet as the "publisher"? 
> If yes, it is almost like many multicast groups being created (in real time?), with subscribers subscribing to a subset of the multicast groups, isn't it? 
> If they are in different subnets, is it like L3 multicast?
> 
> Linda
> 
>> -----Original Message-----
>> From: armd-bounces@ietf.org [mailto:armd-bounces@ietf.org] On Behalf Of
>> Aldrin Isaac
>> Sent: Wednesday, February 15, 2012 8:40 AM
>> To: Thomas Narten
>> Cc: armd@ietf.org
>> Subject: Re: [armd] Multicast in the data center [was Re: address
>> resolution requirement from hosts to overlay edge nodes. Any opinion?]
>> 
>> Any DC with applications performing heavy pub-sub are more than likely
>> using multicast.  I am aware of many companies for whom multicast is
>> critical and enabled on their DC routers.
>> 
>> On Feb 15, 2012, at 9:07 AM, Thomas Narten wrote:
>> 
>>> Mike McBride <mmcbride7@gmail.com> writes:
>>> 
>>>> Large L2 overlay networks?  I don't know. Would be good to find out
>>>> from the community about performance and scalability of multicast in
>>>> the DC.
>>> 
>>> My impression is that many data centers do not enable IP multicast on
>>> their routers. That means you can use link-local multicast (which
>>> works fine within one IP subnet and doesn't really have scaling
>>> issues). But if you want multicast that goes beyond one link (and IP
>>> subnet), which is presumably necessary for an overlay like
>>> VXLAN/NVGRE, that is where you have problems.
>>> 
>>> The question is not even whether L3 multicast scales. It's whether
>> the
>>> DC operater is willing to enable such multicast.
>>> 
>>> Thomas
>>> 
>>> _______________________________________________
>>> armd mailing list
>>> armd@ietf.org
>>> https://www.ietf.org/mailman/listinfo/armd
>> 
>> _______________________________________________
>> armd mailing list
>> armd@ietf.org
>> https://www.ietf.org/mailman/listinfo/armd