Hello, Anyone with opinions on what restrictions a service provider should and should not impose on Ethernet L2 circuits provided to business customers wanting to connect several offices? The service provider's MPLS core network doesn't mind what traffic flows through the EoMPLS tunnel, but the L2 access network do mind and can be vulnerable to several layer 2 issues. Broadcast storm control and BPDU filter will protect the access network to a certain degree, but there are still potential layer 2 problems that can affect the switches, for example MAC address spoofing/flooding. Not to mention potentially difficult troubleshooting if a business customer connects two large LANs through the ISP's Ethernet layer 2 circuit, and then experience some occult layer 2 problem. Should business customers expect to be able to connect several LANs through an Ethernet L2 ciruit and build a layer 2 network spanning several locations? Or should the service provider implement port security and limit the number of MAC addresses on the access ports, forcing the customer to connect a router in both ends and segment their network? Also, do you see a demand for multi-point layer 2 networks (requiring VPLS), or are point-to-point layer 2 circuits sufficient to meet market demand? The most important argument for customers that choose Ethernet L2 over MPLS IP-VPN is that they want full control over their routing, they don't want the involvement from the service provider. Some customers also argue that a flat layer 2 network spanning several locations is a simpler and better design for them, and they don't want the "hassle" with routers and network segmentation. But IMO the customer (and the service provider) is far better off by segmenting their network in the vast majority of cases. What do you think? Regards, Even ___________________________________________________ This message and any attachment is intended for the person(s) named above only. It may contain information that is confidential or legally privileged. If you have received this communication in error, please erase all copies of the message and its attachments and notify us immediately. Thank You. This footnote confirms that the email and attachment(s) has been swept by our anti-virus solution for the presence of known computer viruses.
----- "Endresen Even" <Even.Endresen@bkk.no> schreef:
Hello,
Anyone with opinions on what restrictions a service provider should and should not impose on Ethernet L2 circuits provided to business customers wanting to connect several offices?
Although different in concept, but somehwat similair in issues regarding L2, this description from the Amsterdam IX can give some pointers as to which issues need or can be adressed to minimize L2 issues. See: http://www.ams-ix.net/specifications-descriptions/ Regards, Nils Kolstein SSCPlus B.V.
Interesting questions. Here are a few thoughts from the perspective of an education/research backbone operator that used to be IP only but has also been offering L2 point-to-point circuits for a few years.
Should business customers expect to be able to connect several LANs through an Ethernet L2 ciruit and build a layer 2 network spanning several locations?
At least for our customers, that is indeed important. The most popular application here is for a customer to connect a remote location to their campus network, and that want to (at least be able to) use any of their existing VLANs at the remote site.
Or should the service provider implement port security and limit the number of MAC addresses on the access ports, forcing the customer to connect a router in both ends and segment their network?
That would make the service less attractive, and also more complex to set up and maintain. For point-to-point service, there is really no reason for the network to care about customers' MAC addresses, VLAN tags and such. As you said, EoMPLS doesn't care. (Ethernet over L2TPv3 shouldn't care either. If I had cost-effective edge routers that did L2TPv3 encapsulation/decapsulation at line rate, I'd switch off MPLS in our core tomorrow.) Couldn't PBB or even Q-in-Q provide that isolation as well, at least for point-to-point services? I must say that I don't personally have much experience with those, because we tend to connect our customers to EoMPLS-capable routers directly.
Also, do you see a demand for multi-point layer 2 networks (requiring VPLS), or are point-to-point layer 2 circuits sufficient to meet market demand?
That's a big question for us right now... we're not sure yet. I'd like to hear others' opinions on this.
The most important argument for customers that choose Ethernet L2 over MPLS IP-VPN is that they want full control over their routing, they don't want the involvement from the service provider. Some customers also argue that a flat layer 2 network spanning several locations is a simpler and better design for them, and they don't want the "hassle" with routers and network segmentation.
I have a good deal of sympathy for customers who think this way. Also from the service provider point of view, I like the simplicity of the offering - basically we're providing an emulation of a very long piece of Ethernet cable. (My worry with multipoint L2 VPNs is that they can't have such a simple service model.)
But IMO the customer (and the service provider) is far better off by segmenting their network in the vast majority of cases. What do you think?
Maybe they already have a segmented network, but don't want to segment it based on geography/topology. As far as I'm concerned, enterprises should just connect their various sites to the Internet independently, and use VPN techniques if and where necessary to provide the illusion of a unified network. In practice, this illusion of a single large LAN (or rather, multiple organization-wide LANs) is very important to the typical enterprise, because so much security policy is enforced based on IP addresses. And the typical enterprise wants a central chokepoint that all traffic must go through, for reasons that might have to do with security, or support costs, or with (illusions of) control. This bridging function required to maintain the illusion of a unified network is something that most enterprises prefer to outsource. I'd hope that at some point, better security mechanisms and/or better VPN technologies make these kinds of VPN services less relevant. Until that happens, there's going to be demand for them. Of course the telcos have known that for eons and provided many generations of expensive and hard-to-use services to address this. Point-to-point Ethernet services are interesting because they are relatively easy to provide for folks like us who only really know IP (and maybe some MPLS). And the more transparent they are, the easier it is for customers to use them. -- Simon.
Or should the service provider implement port security and limit the number of MAC addresses on the access ports, forcing the customer to connect a router in both ends and segment their network?
That would make the service less attractive, and also more complex to set up and maintain. For point-to-point service, there is really no reason for the network to care about customers' MAC addresses, VLAN tags and such.
*If* the customer connects directly to a router which terminates EoMPLS, I agree. But router ports are usually expensive, which often means that the customer connects to a switch. And switches definitely care about MAC addresses.
Couldn't PBB or even Q-in-Q provide that isolation as well, at least for point-to-point services? I must say that I don't personally have much experience with those, because we tend to connect our customers to EoMPLS-capable routers directly.
QinQ does nothing to reduce the number of MAC addresses required. PBB can do this, but there is still not a lot of PBB equipment available.
Also, do you see a demand for multi-point layer 2 networks (requiring VPLS), or are point-to-point layer 2 circuits sufficient to meet market demand?
That's a big question for us right now... we're not sure yet. I'd like to hear others' opinions on this.
There is some demand there. Whether that makes it worth it implementing as a product is another question. Trouybleshooting multipoint is more difficult than troubleshooting point to point circuits. Steinar Haug, Nethelp consulting, sthaug@nethelp.no
Couldn't PBB or even Q-in-Q provide that isolation as well, at least for point-to-point services? I must say that I don't personally have much experience with those, because we tend to connect our customers to EoMPLS-capable routers directly.
QinQ does nothing to reduce the number of MAC addresses required. PBB can do this, but there is still not a lot of PBB equipment available.
PBB would help a lot, but as far as Cisco equipment is concerned, it is only supported on 7600 with ES40 line cards: http://www.cisco.com/en/US/docs/ios/cether/configuration/guide/ce_mac_evc_pb... On my wish list is PBB support on access layer switches, like the Cisco ME3400. This would bring the benefits of tunneling out to the very edge of the network. We have an MPLS core with a hierarchial Ethernet layer 2 access network between the core and the customer. Typically there are two or three switches between the the PE node and the customer. Even though we are building the MPLS network further out towards the edge, there will always be a layer 2 access network. It is the switches in the access network that is my concern. We have seen some rather strange problems over the years, like customer nodes that flood MAC address tables with spoofed MAC addresses, and frames that are reflected, causing switches to continually re-learn the same MAC-addresses from different ports. MAC header encapsulation at the very edge of the network would protect the switches in the access layer, and thereby make the service providers more willing to offer more transparent products to their customers. Regards, Even ___________________________________________________ This message and any attachment is intended for the person(s) named above only. It may contain information that is confidential or legally privileged. If you have received this communication in error, please erase all copies of the message and its attachments and notify us immediately. Thank You. This footnote confirms that the email and attachment(s) has been swept by our anti-virus solution for the presence of known computer viruses.
-----Original Message----- From: Simon Leinen Sent: Thursday, December 31, 2009 8:29 AM Subject: Re: Restrictions on Ethernet L2 circuits?
Should business customers expect to be able to connect several LANs through an Ethernet L2 ciruit and build a layer 2 network spanning several locations?
At least for our customers, that is indeed important. The most popular application here is for a customer to connect a remote location to their campus network, and that want to (at least be able to) use any of their existing VLANs at the remote site.
That is what I currently expect from a layer 2 solution. One that does not allow VLAN tagging across the span is of much less utility.
Or should the service provider implement port security and limit the number of MAC addresses on the access ports, forcing the customer to connect a router in both ends and segment their network?
That would make the service less attractive, and also more complex to set up and maintain. For point-to-point service, there is really no reason for the network to care about customers' MAC addresses, VLAN tags and such. As you said, EoMPLS doesn't care. (Ethernet over L2TPv3 shouldn't care either. If I had cost-effective edge routers that did L2TPv3 encapsulation/decapsulation at line rate, I'd switch off MPLS in our core tomorrow.)
Also, do you see a demand for multi-point layer 2 networks (requiring VPLS), or are point-to-point layer 2 circuits sufficient to meet market demand?
That's a big question for us right now... we're not sure yet. I'd
I don't want my provider enforcing such things on me provided it doesn't blow up their network. If I break my stuff, I expect to own all the pieces. I don't want them to nanny me and enforce policy that they determine is "for my own good" but are of no consequence in maintaining their own network. I want the pipe to be basically a long ethernet cable. My traffic should be sufficiently isolated as not to cause a problem in their network no matter what I do. like
to hear others' opinions on this.
I once had such a solution in a network and it worked quite well. It was the (now defunct) Yipes! NAN (National Area Network) product. We used it for OSPF area 0 between all of our US facilities (several offices and two production colos). It worked quite well for the amount of traffic that went between facilities and it was stable for the approx. two years we had it deployed. In that case we had only a single VLAN that acted as a flat layer2 network that spanned the country with a pair of layer 2/3 switches at each office acting as routers for each facility area. This solution turned out to be cheaper and more efficient than running dedicated links to each facility and getting everything meshed over point-to-point circuts. If someone else already has a nationwide mesh, it probably makes good sense to lease some of that capacity than to try and build your own.
-----Original Message----- From: Shane Ronan Sent: Thursday, December 31, 2009 12:24 PM Subject: Re: Restrictions on Ethernet L2 circuits?
(now defunct) Yipes! NAN (National Area Network) product
They don't offer this anymore?
Yipes! Doesn't exist anymore. They were taken over by Reliance Globalcom and I am not sure what their offerings are. I would expect them to offer something similar but possibly under a different name.
Yipes is still offering services under the Yipes, name, at least in the NY Metro Area. On Dec 31, 2009, at 3:32 PM, George Bonser wrote:
-----Original Message----- From: Shane Ronan Sent: Thursday, December 31, 2009 12:24 PM Subject: Re: Restrictions on Ethernet L2 circuits?
(now defunct) Yipes! NAN (National Area Network) product
They don't offer this anymore?
Yipes! Doesn't exist anymore. They were taken over by Reliance Globalcom and I am not sure what their offerings are. I would expect them to offer something similar but possibly under a different name.
-----Original Message----- From: Shane Ronan Sent: Thursday, December 31, 2009 1:39 PM Subject: Re: Restrictions on Ethernet L2 circuits?
Yipes is still offering services under the Yipes, name, at least in
the
NY Metro Area.
I based my information on the fact that going to yipes.com takes you to Reliance. And this "The Yipes name is (presumably) gone forever, as the company has been assimilated into a new global communications entity, Reliance Globalcom. Virtually unknown in the U.S., Reliance will need to spend heavily on branding and marketing efforts." And this: Yipes was acquired by Reliance Globalcom in December 2007. Reliance Globalcom, a division of Reliance Communications, spearheads the Global Telecom operations of India's largest Integrated Telecom Service Provider ... So while they might be operating still as Yipes in some areas, they are Reliance Globalcom out here in California.
The MEF has a set of specs for this. http://metroethernetforum.org/ In general, it's built as a "dumb pipe" virtual circuit, IE your client BPDUs and other IEEE 802.* signaling are ignored, as they are encapsulated, and forwarded explicitly to a given port. What you do on the switch that gets the deencapsualted traffic is your business. -----Original Message----- From: Endresen Even [mailto:Even.Endresen@bkk.no] Sent: Thursday, December 31, 2009 12:41 AM To: nanog@nanog.org Subject: Restrictions on Ethernet L2 circuits? Hello, Anyone with opinions on what restrictions a service provider should and should not impose on Ethernet L2 circuits provided to business customers wanting to connect several offices? The service provider's MPLS core network doesn't mind what traffic flows through the EoMPLS tunnel, but the L2 access network do mind and can be vulnerable to several layer 2 issues. Broadcast storm control and BPDU filter will protect the access network to a certain degree, but there are still potential layer 2 problems that can affect the switches, for example MAC address spoofing/flooding. Not to mention potentially difficult troubleshooting if a business customer connects two large LANs through the ISP's Ethernet layer 2 circuit, and then experience some occult layer 2 problem. Should business customers expect to be able to connect several LANs through an Ethernet L2 ciruit and build a layer 2 network spanning several locations? Or should the service provider implement port security and limit the number of MAC addresses on the access ports, forcing the customer to connect a router in both ends and segment their network? Also, do you see a demand for multi-point layer 2 networks (requiring VPLS), or are point-to-point layer 2 circuits sufficient to meet market demand? The most important argument for customers that choose Ethernet L2 over MPLS IP-VPN is that they want full control over their routing, they don't want the involvement from the service provider. Some customers also argue that a flat layer 2 network spanning several locations is a simpler and better design for them, and they don't want the "hassle" with routers and network segmentation. But IMO the customer (and the service provider) is far better off by segmenting their network in the vast majority of cases. What do you think? Regards, Even ___________________________________________________ This message and any attachment is intended for the person(s) named above only. It may contain information that is confidential or legally privileged. If you have received this communication in error, please erase all copies of the message and its attachments and notify us immediately. Thank You. This footnote confirms that the email and attachment(s) has been swept by our anti-virus solution for the presence of known computer viruses.
-----Original Message----- From: Endresen Even Sent: Thursday, December 31, 2009 12:41 AM Subject: Restrictions on Ethernet L2 circuits?
Hello,
Anyone with opinions on what restrictions a service provider should and should not impose on Ethernet L2 circuits provided to business customers wanting to connect several offices?
One thing I *really* miss from about a decade back is the Telseon model. It was completely self-managed. They would place a switch at the customer's location and if I wanted to change the bandwidth allocation on a port, it was as easy as logging in to a management portal and changing it. So if a port usually required only 10Meg but I needed to do something different for a few days, I could bump the bandwidth of the port up to 100Meg and then set it back down when I was done and I was billed based on what each day's bandwidth setting was. We paid only for what we had configured for each day. The other benefit of it was that if someone else also had the same service, we could self-provision a "logical wire" between us. So if I wanted to peer or somehow partner with another network that also had a Telseon switch, we simply logged in to the management portal and configured it, jacked in to the appropriate port on our respective switches, and it was done. It didn't even take a phone call, the provisioning process could be done on the web. I don't think anyone offers such a MAN service these days and I really wish it had stayed around.
participants (7)
-
Endresen Even -
George Bonser -
Nils Kolstein -
Shane Ronan -
Simon Leinen -
sthaug＠nethelp.no -
Tomas L. Byrnes