Login | Register For Free | Help
Search for: (Advanced)

Mailing List Archive: NANOG: users

Rate shaping in Active E FTTx networks

 

 

NANOG users RSS feed   Index | Next | Previous | View Threaded


jason at lixfeld

Jul 26, 2012, 12:45 PM

Post #1 of 7 (414 views)
Permalink
Rate shaping in Active E FTTx networks

Hi all,

I'm trying to gauge what operators are doing to handle per-subscriber Internet access PIR bandwidth in Active E FTTx networks.

I presume operators would want to limit the each subscriber to a certain PIR, but within that limit, do things like perform preferential treatment of interactive services like steaming video or Skype, etc., ahead of non-interactive services like FTP.

My impression is that a subscriber's physical access in these networks is exponentially larger than their allocated amount of Internet access. This would leave ample room on the physical access access for other services like Voice and IPTV that might run on separate VLANs than the Internet access VLAN. That said, I doubt there's really that much of a concern about allocating PIR on these other service VLANs.

So in terms of PIR for Internet access, is there some magic box that sits between the various subscriber aggregation points and the core, which takes care of shaping the subscriber's Internet access PIR, while making sure that the any preferential treatment of interactive services is performed.

Is that a lot to ask for one box? The ridiculously deep buffers required in order to shape to PIR vs. police to it (because policing to a PIR is just plain ugly) and the requirements to perform any sort of preferential packet treatment above and beyond that seem like quite a lot to ask of one box. Am I wrong?

Who might make a box like this, if it exists? And if not, what are folks using the achieve these results?

Thanks in advance for any insights..


erikm at buh

Jul 26, 2012, 8:21 PM

Post #2 of 7 (373 views)
Permalink
Re: Rate shaping in Active E FTTx networks [In reply to]

On 7/26/12 12:45 , Jason Lixfeld wrote:
> Hi all,
>
> I'm trying to gauge what operators are doing to handle per-subscriber
> Internet access PIR bandwidth in Active E FTTx networks.
>
> I presume operators would want to limit the each subscriber to a
> certain PIR, but within that limit, do things like perform preferential
> treatment of interactive services like steaming video or Skype, etc.,
> ahead of non-interactive services like FTP.
>
> My impression is that a subscriber's physical access in these networks
> is exponentially larger than their allocated amount of Internet access.
> This would leave ample room on the physical access access for other
> services like Voice and IPTV that might run on separate VLANs than the
> Internet access VLAN. That said, I doubt there's really that much of a
> concern about allocating PIR on these other service VLANs.
>
> So in terms of PIR for Internet access, is there some magic box that
> sits between the various subscriber aggregation points and the core,
> which takes care of shaping the subscriber's Internet access PIR, while
> making sure that the any preferential treatment of interactive services
> is performed.
>
> Is that a lot to ask for one box? The ridiculously deep buffers
> required in order to shape to PIR vs. police to it (because policing to
> a PIR is just plain ugly) and the requirements to perform any sort of
> preferential packet treatment above and beyond that seem like quite a
> lot to ask of one box. Am I wrong?
>
> Who might make a box like this, if it exists? And if not, what are
> folks using the achieve these results?
>
> Thanks in advance for any insights..

I've seen a few deployments using Packeteer's (now BlueCoat) PacketShaper
for this purpose; the only downside I've heard with that platform is cost.
Sandvine and Fortinet are a couple other options that have different
approaches, but have a lot of this functionality rolled in alongside their
broader security services.

-e


MGauvin at dryden

Jul 26, 2012, 8:48 PM

Post #3 of 7 (376 views)
Permalink
RE: Rate shaping in Active E FTTx networks [In reply to]

Juniper dynamic application awareness does a decent job and so does the cisco counterpart

saves buying more hw
________________________________________
From: Erik Muller [erikm [at] buh]
Sent: Thursday, July 26, 2012 10:21 PM
To: nanog [at] nanog
Subject: Re: Rate shaping in Active E FTTx networks

On 7/26/12 12:45 , Jason Lixfeld wrote:
> Hi all,
>
> I'm trying to gauge what operators are doing to handle per-subscriber
> Internet access PIR bandwidth in Active E FTTx networks.
>
> I presume operators would want to limit the each subscriber to a
> certain PIR, but within that limit, do things like perform preferential
> treatment of interactive services like steaming video or Skype, etc.,
> ahead of non-interactive services like FTP.
>
> My impression is that a subscriber's physical access in these networks
> is exponentially larger than their allocated amount of Internet access.
> This would leave ample room on the physical access access for other
> services like Voice and IPTV that might run on separate VLANs than the
> Internet access VLAN. That said, I doubt there's really that much of a
> concern about allocating PIR on these other service VLANs.
>
> So in terms of PIR for Internet access, is there some magic box that
> sits between the various subscriber aggregation points and the core,
> which takes care of shaping the subscriber's Internet access PIR, while
> making sure that the any preferential treatment of interactive services
> is performed.
>
> Is that a lot to ask for one box? The ridiculously deep buffers
> required in order to shape to PIR vs. police to it (because policing to
> a PIR is just plain ugly) and the requirements to perform any sort of
> preferential packet treatment above and beyond that seem like quite a
> lot to ask of one box. Am I wrong?
>
> Who might make a box like this, if it exists? And if not, what are
> folks using the achieve these results?
>
> Thanks in advance for any insights..

I've seen a few deployments using Packeteer's (now BlueCoat) PacketShaper
for this purpose; the only downside I've heard with that platform is cost.
Sandvine and Fortinet are a couple other options that have different
approaches, but have a lot of this functionality rolled in alongside their
broader security services.

-e


jeff-kell at utc

Jul 27, 2012, 5:01 AM

Post #4 of 7 (379 views)
Permalink
Re: Rate shaping in Active E FTTx networks [In reply to]

On 7/26/2012 11:21 PM, Erik Muller wrote:
> I've seen a few deployments using Packeteer's (now BlueCoat)
> PacketShaper for this purpose; the only downside I've heard with that
> platform is cost. Sandvine and Fortinet are a couple other options
> that have different approaches, but have a lot of this functionality
> rolled in alongside their broader security services.

For shaping flexibility and real DPI, Procera PacketLogic is an order of
magnitude (and throughput) beyond Packeteer (speaking as a current user
of the former and a former user of the latter). I know their higher-ed
distribution is substantial (for those that shape by policy). There are
other "fair game" shaping appliances (NetEqualizer) if you just want to
give everyone equal access to whatever bandwidth remains. But for real
application inspection, the traditional players (Packeteer, Allot, etc)
today just tell you that yes, 80-90% of your traffic is HTTP protocol,
now what?

Jeff


jared at puck

Jul 27, 2012, 5:16 AM

Post #5 of 7 (374 views)
Permalink
Re: Rate shaping in Active E FTTx networks [In reply to]

Many CPE platforms have the rate limit built in. Some (eg: Zhone) do this in 1mbps increments. Ideally there would be some greater level of granularity but it seems to work. You can obviously police on the other end as well if required.

Jared Mauch

On Jul 26, 2012, at 3:45 PM, Jason Lixfeld <jason [at] lixfeld> wrote:

> So in terms of PIR for Internet access, is there some magic box that sits between the various subscriber aggregation points and the core, which takes care of shaping the subscriber's Internet access PIR, while making sure that the any preferential treatment of interactive services is performed.


bedard.phil at gmail

Jul 27, 2012, 5:49 AM

Post #6 of 7 (376 views)
Permalink
Re: Rate shaping in Active E FTTx networks [In reply to]

On the downstream end the limiting is usually done on the subscriber aggregation equipment. Router vendors sell linecards with large amounts of queue capability for this reason. This is where you would introduce some kind of QoS to deal with video or voice as well. Upstream could be done the same way if they have true direct connections to the gear or be done on a CPE.

As far as differentiating traffic within an Internet pipe that is a slippery legal slope. Others have mentioned the bigger players like Procera and Sandvine.

Phil

On Jul 26, 2012, at 3:45 PM, Jason Lixfeld <jason [at] lixfeld> wrote:

> Hi all,
>
> I'm trying to gauge what operators are doing to handle per-subscriber Internet access PIR bandwidth in Active E FTTx networks.
>
> I presume operators would want to limit the each subscriber to a certain PIR, but within that limit, do things like perform preferential treatment of interactive services like steaming video or Skype, etc., ahead of non-interactive services like FTP.
>
> My impression is that a subscriber's physical access in these networks is exponentially larger than their allocated amount of Internet access. This would leave ample room on the physical access access for other services like Voice and IPTV that might run on separate VLANs than the Internet access VLAN. That said, I doubt there's really that much of a concern about allocating PIR on these other service VLANs.
>
> So in terms of PIR for Internet access, is there some magic box that sits between the various subscriber aggregation points and the core, which takes care of shaping the subscriber's Internet access PIR, while making sure that the any preferential treatment of interactive services is performed.
>
> Is that a lot to ask for one box? The ridiculously deep buffers required in order to shape to PIR vs. police to it (because policing to a PIR is just plain ugly) and the requirements to perform any sort of preferential packet treatment above and beyond that seem like quite a lot to ask of one box. Am I wrong?
>
> Who might make a box like this, if it exists? And if not, what are folks using the achieve these results?
>
> Thanks in advance for any insights..


mark.tinka at seacom

Jul 29, 2012, 7:12 AM

Post #7 of 7 (359 views)
Permalink
Re: Rate shaping in Active E FTTx networks [In reply to]

On Thursday, July 26, 2012 09:45:14 PM Jason Lixfeld wrote:

> Is that a lot to ask for one box? The ridiculously deep
> buffers required in order to shape to PIR vs. police to
> it (because policing to a PIR is just plain ugly) and
> the requirements to perform any sort of preferential
> packet treatment above and beyond that seem like quite a
> lot to ask of one box. Am I wrong?

Having used middleware in the past to do bandwidth
management, this doesn't scale well when your network grows,
and when off-net traffic (including that between your own
customers) is coming in from several points in the backbone.

On smaller networks, having middleware is easy because your
exit points are finite and fairly static. When you grow and
start peering, taking on several large customers that want
to talk to each other across your network, middleware
becomes cumbersome to deploy, because then not only can't
you assume that 80% of your traffic is HTTP, but you also
can't assume that 80% of your traffic is toward your
upstreams. Moreover, adding redundancy (as in multiple links
between routers/switches) makes the situation worse, because
middleware might not be as inclined, and arbitration of
bandwidth management across multiple middleware devices to
avoid accidentally over-provisioning to customers gets
expensive and complex.

I've since migrated to performing bandwidth management in
the router gear itself. This is easy if you're using high-
end kit (think Juniper M/MX/T, Cisco ASR1000/9000), but
significantly less so on wireline Metro-E networks (where
your Active-E comes in). But not anymore - there have been
meaningful developments in this area, and for some I've had
the pleasure of deploying, e.g., Cisco's ME3600X/3800X.
There also alternatives like Juniper's MX80 (too big, I
think, but the smallest you can get from them now) and
Brocade's NetIron CES/CER2000 units. These allow you to not
only gain decent feature set in the Access, but also let you
extend IP (and MPLS) into the edge too for additional
simplicity.

Hope this helps.

Cheers

Mark.
Attachments: signature.asc (0.82 KB)

NANOG users RSS feed   Index | Next | Previous | View Threaded
 
 


Interested in having your list archived? Contact Gossamer Threads
 
  Web Applications & Managed Hosting Powered by Gossamer Threads Inc.