Re: Using /31 for router links
On 23/01/10 19:52, Michael Sokolov wrote: Mark Smith na...@85d5b20a518b8f6864949bd940457dc124746ddc.nosense.org wrote: As for ATM... The part that totally baffles me about the use of ATM on xDSL lines is that I have never, ever, ever seen an xDSL line carrying more than one ATM VC. OK, there may be someone out there who has set up a configuration like that just for fun, but 99.999% of all ATM'd xDSL lines out there carry a single PVC at 0*35 or 0*38. It's common practice down here in Italy to have more than one VC. One is used to carry data and the other one is used for VoIP, so that you don't have to do QoS on the data part. Ciao ! -- Massimiliano Stucchi BrianTel Srl stuc...@briantel.com Tel (+39) 039 9669921 | Fax (+39) 02 44417204 I-23807, Merate (Lecco), via Mameli 6 MS16801-RIPE
RE: Using /31 for router links
We use 5 PVCs for the IP video and one for Internet. Not as uncommon as you think. Frank -Original Message- From: Michael Sokolov [mailto:msoko...@ivan.harhan.org] Sent: Saturday, January 23, 2010 12:53 PM To: nanog@nanog.org Subject: Re: Using /31 for router links Mark Smith na...@85d5b20a518b8f6864949bd940457dc124746ddc.nosense.org wrote: What about NAT, ATM cell tax, unnecessary addressing fields in PTP protocols (including your beloved HDLC), SSAP, DSAP fields not being big enough in 802.2 necessitating SNAP, IPX directly over 802.3, AAL1 through AAL4, PPPoE dumbell MTUs and MSS hacks? Some of those are far worse sins in my opinion. snip As for ATM... The part that totally baffles me about the use of ATM on xDSL lines is that I have never, ever, ever seen an xDSL line carrying more than one ATM VC. OK, there may be someone out there who has set up a configuration like that just for fun, but 99.999% of all ATM'd xDSL lines out there carry a single PVC at 0*35 or 0*38. So what then is the point of running ATM?!?! All the hyped benefits of ATM (a little cell can squeeze in the middle of a big packet without waiting for it to finish, yadda yadda yadda) are contingent upon having more than one VPI/VCI going across the interface! If every single non-idle cell going across that ATM interface is 0*35 or 0*38, the interface will never carry anything other a direct succession of cells making up an AAL5 packet, strictly in sequence and without interruption. So what's the point of ATM then? Why chop that packet up into cells only to transmit those cells in direct sequence one after another? Why not simply send that same packet in plain HDLC over the same copper pairs/fiber? OK, the backhaul network upstream of the DSLAM may be ATM and that one does have many VCs, so ATM *might* be of use there, but even in that case why not do FRF.8 in the DSLAM and keep the ATM strictly on the backhaul, sending HDLC down the copper pairs? off the soapbox for the moment MS
Re: Using /31 for router links
* Seth Mattinen: In the past I've always used /30's for PTP connection subnets out of old habit (i.e. Ethernet that won't take unnumbered) but now I'm considering switching to /31's in order to stretch my IPv4 space further. Has anyone else does this? Good? Bad? Bad. For some systems, such tricks work to some degree only due to lack of input validation, and you get failures down the road (ARP ceases to work, packet filters are not applied properly and other fun). And now is not the time to conserve address space. You really should do everything you can to justify additional allocations from your RIR.
Re: Using /31 for router links
That's a vendor specific issue. Maybe you could take it up with them and ask what year they think this is? tv - Original Message - From: Florian Weimer f...@deneb.enyo.de To: Seth Mattinen se...@rollernet.us Cc: nanOG list nanog@nanog.org Sent: Saturday, January 23, 2010 4:06 AM Subject: Re: Using /31 for router links * Seth Mattinen: In the past I've always used /30's for PTP connection subnets out of old habit (i.e. Ethernet that won't take unnumbered) but now I'm considering switching to /31's in order to stretch my IPv4 space further. Has anyone else does this? Good? Bad? Bad. For some systems, such tricks work to some degree only due to lack of input validation, and you get failures down the road (ARP ceases to work, packet filters are not applied properly and other fun). And now is not the time to conserve address space. You really should do everything you can to justify additional allocations from your RIR.
Re: Using /31 for router links
Mark Smith na...@85d5b20a518b8f6864949bd940457dc124746ddc.nosense.org wrote: What about NAT, ATM cell tax, unnecessary addressing fields in PTP protocols (including your beloved HDLC), SSAP, DSAP fields not being big enough in 802.2 necessitating SNAP, IPX directly over 802.3, AAL1 through AAL4, PPPoE dumbell MTUs and MSS hacks? Some of those are far worse sins in my opinion. Hmm. PPPoE: this kludge is a direct fallout of abusing Ethernet for WAN/PTP. If all those xDSL users were willing to stick V.35 cards in their PCs and use modems that put out V.35 instead of Ethernet, the whole PPPoE kludge with all of its attendant MTU issues would have been completely unnecessary. Want PPP for authentication etc? Just run straight PPP (RFC 1662) over V.35 instead of Ethernet/PPPoE, HDLC has no fixed MTU unlike Ethernet (jogging my memory, all HDLC controllers which I recall working with allowed maximum frame size up to just a little under 2^16 octets or so), and one can thus have the standard MTU of 1500 octets on that PPP link! Oh, and yet another soapbox of mine, an xDSL modem that puts out V.35 instead of goddamn Ethernet would be a true modem: a modulator/demodulator that modulates/demodulates the bits at the electrical level without caring about what's in those bits. What everyone else in this fubared world calls an xDSL modem (a black box that puts Ethernet out) is not a modem at all (i.e., total misappropriation of the term), it is actually a bastardized router! These boxes forward packets between two network interfaces: the presented Ethernet interface and the internal (often horrendously non-standard and proprietary) HDLC or ATM interface on the actual line. A device that forwards packets between two different network interfaces is by definition a router, hence what everyone calls a modem is actually a bastardized router - bastardized because its routing (packet forwarding) function is something incomprehensible. The Ethernet-to-Ethernet NAT boxes that everyone else calls routers should be called NATters or something like that, anything but a router! A true router is a box with a few AUIs and a few V.35 ports sticking out of it, running some very capable, flexible and totally user-configurable packet forwarding software stack that supports all networking models: IP routing, MAC bridging, VC cross-connect. As for ATM... The part that totally baffles me about the use of ATM on xDSL lines is that I have never, ever, ever seen an xDSL line carrying more than one ATM VC. OK, there may be someone out there who has set up a configuration like that just for fun, but 99.999% of all ATM'd xDSL lines out there carry a single PVC at 0*35 or 0*38. So what then is the point of running ATM?!?! All the hyped benefits of ATM (a little cell can squeeze in the middle of a big packet without waiting for it to finish, yadda yadda yadda) are contingent upon having more than one VPI/VCI going across the interface! If every single non-idle cell going across that ATM interface is 0*35 or 0*38, the interface will never carry anything other a direct succession of cells making up an AAL5 packet, strictly in sequence and without interruption. So what's the point of ATM then? Why chop that packet up into cells only to transmit those cells in direct sequence one after another? Why not simply send that same packet in plain HDLC over the same copper pairs/fiber? OK, the backhaul network upstream of the DSLAM may be ATM and that one does have many VCs, so ATM *might* be of use there, but even in that case why not do FRF.8 in the DSLAM and keep the ATM strictly on the backhaul, sending HDLC down the copper pairs? off the soapbox for the moment MS
Re: Using /31 for router links
Florian Weimer f...@deneb.enyo.de writes: Bad. For some systems, such tricks work to some degree only due to lack of input validation, and you get failures down the road (ARP ceases to work, packet filters are not applied properly and other fun). I never had any problems using Cisco to Cisco, Linux to Linux or Cisco to Linux using /31. Only problem I encountered was when a Linux based router was replaced by a Windows box (please don't ask). cheers Jens -- - | Foelderichstr. 40 | 13595 Berlin, Germany | +49-151-18721264 | | http://www.quux.de | http://blog.quux.de | jabber: jensl...@guug.de | -
Re: Using /31 for router links
Chris Costa cco...@cenic.org writes: We recently did a backbone router upgrade and the vendor surprisingly didn't support /31's. Mind dropping a name? Jens -- - | Foelderichstr. 40 | 13595 Berlin, Germany | +49-151-18721264 | | http://www.quux.de | http://blog.quux.de | jabber: jensl...@guug.de | -
RE: Using /31 for router links
As for ATM... The part that totally baffles me about the use of ATM on xDSL lines is that I have never, ever, ever seen an xDSL line carrying more than one ATM VC. OK, there may be someone out there who has set up a configuration like that just for fun, but 99.999% of all ATM'd xDSL lines out there carry a single PVC at 0*35 or 0*38. Multi-PVC is used (in the context of xSLAM--CPE), for example, for delivering IPTV+DSL. 0/35 and 0/38 are just arbitrary numbers, there are plenty of other random ones like 0/33 used by major service providers. Arguably your 99.999% is way off.
Re: Using /31 for router links
[Michael Sokolov said:] *snip* but 99.999% of all ATM'd xDSL lines out there carry a single PVC at 0*35 or 0*38. So what then is the point of running ATM?!?! *snip* We've got several ADSL and SDSL circuits that carry two PVC's: 0/35 and 0/36. Covad has a product called Voice Optimized Access. I won't go into the gory details of the underlying technology, but the second PVC is utilized for voice. Essentially two separate IP networks over one physical network, one with guaranteed bandwidth (by way of vbr-rt) and QoS through Covad's network. Sincerely, Bobby Glover Director of Information Services South Valley Internet -Original Message- From: msoko...@ivan.harhan.org (Michael Sokolov) Date: Sat, 23 Jan 2010 18:52:51 To: nanog@nanog.org Subject: Re: Using /31 for router links Mark Smith na...@85d5b20a518b8f6864949bd940457dc124746ddc.nosense.org wrote: What about NAT, ATM cell tax, unnecessary addressing fields in PTP protocols (including your beloved HDLC), SSAP, DSAP fields not being big enough in 802.2 necessitating SNAP, IPX directly over 802.3, AAL1 through AAL4, PPPoE dumbell MTUs and MSS hacks? Some of those are far worse sins in my opinion. Hmm. PPPoE: this kludge is a direct fallout of abusing Ethernet for WAN/PTP. If all those xDSL users were willing to stick V.35 cards in their PCs and use modems that put out V.35 instead of Ethernet, the whole PPPoE kludge with all of its attendant MTU issues would have been completely unnecessary. Want PPP for authentication etc? Just run straight PPP (RFC 1662) over V.35 instead of Ethernet/PPPoE, HDLC has no fixed MTU unlike Ethernet (jogging my memory, all HDLC controllers which I recall working with allowed maximum frame size up to just a little under 2^16 octets or so), and one can thus have the standard MTU of 1500 octets on that PPP link! Oh, and yet another soapbox of mine, an xDSL modem that puts out V.35 instead of goddamn Ethernet would be a true modem: a modulator/demodulator that modulates/demodulates the bits at the electrical level without caring about what's in those bits. What everyone else in this fubared world calls an xDSL modem (a black box that puts Ethernet out) is not a modem at all (i.e., total misappropriation of the term), it is actually a bastardized router! These boxes forward packets between two network interfaces: the presented Ethernet interface and the internal (often horrendously non-standard and proprietary) HDLC or ATM interface on the actual line. A device that forwards packets between two different network interfaces is by definition a router, hence what everyone calls a modem is actually a bastardized router - bastardized because its routing (packet forwarding) function is something incomprehensible. The Ethernet-to-Ethernet NAT boxes that everyone else calls routers should be called NATters or something like that, anything but a router! A true router is a box with a few AUIs and a few V.35 ports sticking out of it, running some very capable, flexible and totally user-configurable packet forwarding software stack that supports all networking models: IP routing, MAC bridging, VC cross-connect. As for ATM... The part that totally baffles me about the use of ATM on xDSL lines is that I have never, ever, ever seen an xDSL line carrying more than one ATM VC. OK, there may be someone out there who has set up a configuration like that just for fun, but 99.999% of all ATM'd xDSL lines out there carry a single PVC at 0*35 or 0*38. So what then is the point of running ATM?!?! All the hyped benefits of ATM (a little cell can squeeze in the middle of a big packet without waiting for it to finish, yadda yadda yadda) are contingent upon having more than one VPI/VCI going across the interface! If every single non-idle cell going across that ATM interface is 0*35 or 0*38, the interface will never carry anything other a direct succession of cells making up an AAL5 packet, strictly in sequence and without interruption. So what's the point of ATM then? Why chop that packet up into cells only to transmit those cells in direct sequence one after another? Why not simply send that same packet in plain HDLC over the same copper pairs/fiber? OK, the backhaul network upstream of the DSLAM may be ATM and that one does have many VCs, so ATM *might* be of use there, but even in that case why not do FRF.8 in the DSLAM and keep the ATM strictly on the backhaul, sending HDLC down the copper pairs? off the soapbox for the moment MS
Re: Using /31 for router links
* Tony Varriale: That's a vendor specific issue. Maybe you could take it up with them and ask what year they think this is? I think they support it on point-to-point media only, which seems sufficient for RFC 3021 compliance. Ethernet is a different story, unfortunately.
Re: Using /31 for router links
Michael Sokolov wrote: That is why I hate Ethernet with a passion. Ethernet should be for LANs only; using Ethernet for WANs and PTP links is the vilest invention in the entire history of data networking in my opinion. Ah, but who's to say that all PTP links are WANs? Are you really going to run an OC-48 from one router to another _in the same building_ when you need 1Gb/s between them? Have you looked at how much more that would cost? Ethernet interfaces, particularly copper, are dirt cheap. Even for MANs or WANs, the price of a pipe (plus equipment at each end) will still often be significantly lower for Ethernet than for real circuits--especially if you don't plan on using all the bandwidth all of the time. My medium of choice for PTP links (WAN) is HDLC over a synchronous serial bit stream, with a V.35 or EIA-530 interface between the router and the modem/DSU. Over HDLC I then run either RFC 1490 routed mode or straight PPP (RFC 1662); in the past I used Cisco HDLC (0F 00 08 00 IP header follows...). My 4.3BSD router (or I should better say gateway as that's the proper 80s/90s term) then sees a PTP interface which has no netmask at all, hence the near and far end IP addresses don't have to have any numerical relationship between them at all. No netmask, no MAC addresses, no ARP, none of that crap, just a PTP IP link. Well, it'd certainly be nice if someone would make something even cheaper than Ethernet for that purpose (which would squeeze out a few more bits of payload), but so far nobody has. It's hard to beat Ethernet on volume, and that's the main determinant of cost/price... S -- Stephen Sprunk God does not play dice. --Albert Einstein CCIE #3723 God is an inveterate gambler, and He throws the K5SSSdice at every possible opportunity. --Stephen Hawking smime.p7s Description: S/MIME Cryptographic Signature
Re: Using /31 for router links
Stephen Sprunk step...@sprunk.org wrote: Ah, but who's to say that all PTP links are WANs? Are you really going to run an OC-48 from one router to another _in the same building_ when you need 1Gb/s between them? Can't say - I have never needed that much bandwidth. :) I still live in an alternate Universe where 10 Mbps coaxial Ethernet for LANs is near- infinity and 2 Mbps or so makes for a *very* sweet WAN. The facility housing the mail server from which I am sending this message is connected to the outside Inet via a 384 kbps SDSL pipe which I am using basically as ARPANET replacement - I miss the ARPANET. If I wanted a PTP link between two routers in the same building that runs at the same speed as my Ethernet (10 Mbps), I would use EIA-422 (which is rated up to 10 Mbps) and run something HDLC-based over it. Even for MANs or WANs, the price of a pipe (plus equipment at each end) will still often be significantly lower for Ethernet than for real circuits Wait a moment here. With a MAN/WAN involving wires/fiber running over public property, what one is paying for is the right to use those wires for your data, right? The wires themselves do NOT run Ethernet at the electrical level, so if you have some MAN/WAN Ethernet service, there is a black box of some kind that converts the native electrical signal format to Ethernet. Why not take that black box out of service, use it for baseball practice (Office Space style), and use the exact same wires/fiber (rented at exactly the same monthly recurring price) in its native non-Ethernet form? IOW, if you are renting dry copper / dark fiber, you have a choice to use it either through a stinky black box Ethernet converter or in the native non-Ethernet form directly, but the monthly recurring cost remains exactly the same. Well, it'd certainly be nice if someone would make something even cheaper than Ethernet for that purpose (which would squeeze out a few more bits of payload), but so far nobody has. It's hard to beat Ethernet on volume, and that's the main determinant of cost/price... But that's non-recurring equipment cost only, and at least in my case the little investment in V.35 etc hardware is a much lower cost than the price of pain and suffering with Ethernet for a purist like me. MS
Re: Using /31 for router links
On 1/23/10 11:52 AM, Michael Sokolov wrote: Oh, and yet another soapbox of mine, an xDSL modem that puts out V.35 instead of goddamn Ethernet would be a true modem: a modulator/demodulator that modulates/demodulates the bits at the electrical level without caring about what's in those bits. Back in the days of Rhythms and Copper Mountain gear, Netopia had the D series routers which were actually xDSL to DSU units. Used to use them for customers who had T1 equipment (2500s, 2600s, 1600s, etc). Worked quite well. Though, I'm not sure they'd work these days, nor do I think they came in ADSL models either. -- Brielle Bruns The Summit Open Source Development Group http://www.sosdg.org/ http://www.ahbl.org
Re: Using /31 for router links
Brielle Bruns br...@2mbit.com wrote: Back in the days of Rhythms and Copper Mountain gear, Netopia had the D series routers which were actually xDSL to DSU units. Yes, I am very familiar with them: http://ifctfvax.Harhan.ORG/OpenSDSL/existing_cpe/netopia/dsu.html As that page explains, they are only pseudo-DSUs though. I much much prefer a true bit-transparent DSU: http://ifctfvax.Harhan.ORG/OpenSDSL/existing_cpe/xsb2000.html I have also designed and built an SDSL to EIA-530 DSU of my very own: http://ifctfvax.Harhan.ORG/OpenSDSL/OSDCU/ On the latter I have the hardware (the page above has a photo of the board), but the operational software (or firmware if you will) remains to be finished. Though, I'm not sure they'd work these days, Only in the very limited geographic footpring of what used to be DSL.net - they are the last remaining semi-major operator of CM DSLAMs: http://ifctfvax.Harhan.ORG/OpenSDSL/megapath.html My big goal is to make my own DSU which I have just mentioned function as a Layer 2 converter (FRF.8 and friends) from Nokia SDSL/ATM served by Covad to HDLC. nor do I think they came in ADSL models either. I don't think anyone have *ever* used V.35 friends with ADSL - probably because those who would want V.35 (i.e., people like me) would find ADSL morally offensive. :-) MS
Re: Using /31 for router links
Michael Sokolov wrote: Wait a moment here. With a MAN/WAN involving wires/fiber running over public property, what one is paying for is the right to use those wires for your data, right? The wires themselves do NOT run Ethernet at the electrical level, so if you have some MAN/WAN Ethernet service, there is a black box of some kind that converts the native electrical signal format to Ethernet. Why not take that black box out of service, use it for baseball practice (Office Space style), and use the exact same wires/fiber (rented at exactly the same monthly recurring price) in its native non-Ethernet form? Well, I have an OC-12 upstairs that has an Overture box attached to it because: Ethernet is not tariffed and the OC-12 is. The price difference between the two was substantial (even though it's the same thing) and I was going for cheap alternate path. If they were the same price I would have probably just taken it directly. ~Seth
Re: Using /31 for router links
Greetings, On Fri, 22 Jan 2010, Seth Mattinen wrote: In the past I've always used /30's for PTP connection subnets out of old habit (i.e. Ethernet that won't take unnumbered) but now I'm considering switching to /31's in order to stretch my IPv4 space further. Has anyone else does this? Good? Bad? Based on the bit of testing I've done this shouldn't be a problem since it's only between routers. Yes, this *IS* done *ALL* the time. P-t-P means that there are ONLY two devices on the wire - hence point to point. It ONLY uses two IP addresses (one on each end) and there is no reason or need to ARP on this wire. So no need for a broadcast or network addresses - it is just the two end points. --- Jay Nugent Nugent Telecommunications Train how you will Operate, and you will Operate how you were Trained. ++ | Jay Nugent j...@nuge.com(734)484-5105(734)649-0850/Cell | | Nugent Telecommunications [www.nuge.com]| | Internet Consulting/Linux SysAdmin/Engineering Design/ISP Reseller | | ISP Monitoring [www.ispmonitor.org] ISP Modem Performance Monitoring | | Web-Pegasus[www.webpegasus.com] Web Hosting/DNS Hosting/Shell Accts| ++ 7:01pm up 43 days, 18:42, 3 users, load average: 1.10, 0.96, 0.63
Re: Using /31 for router links
We recently did a backbone router upgrade and the vendor surprisingly didn't support /31's. We had to renumber all those interconnects and peering sessions to /30's. That wasn't fun! On Jan 22, 2010, at 4:53 PM, Seth Mattinen wrote: Joe Provo wrote: On Fri, Jan 22, 2010 at 04:08:28PM -0800, Seth Mattinen wrote: In the past I've always used /30's for PTP connection subnets out of old habit (i.e. Ethernet that won't take unnumbered) but now I'm considering switching to /31's in order to stretch my IPv4 space further. Has anyone else does this? Good? Bad? Based on the bit of testing I've done this shouldn't be a problem since it's only between routers. rfc3021 is over 9 years old, so should be no suprise that it works well. :-) I'm never surprised anymore by something that should work turning out to have some obscure quirk about it, so I figured it was worth asking. ;) ~Seth
RE: Using /31 for router links
rfc3021 is over 9 years old, so should be no suprise that it works well. :-) I'm never surprised anymore by something that should work turning out to have some obscure quirk about it, so I figured it was worth asking. ;) It's not a quirk, it's an implementation-specific feature ;)
Re: Using /31 for router links
On Jan 22, 2010, at 4:41 PM, Joe Provo wrote: On Fri, Jan 22, 2010 at 04:08:28PM -0800, Seth Mattinen wrote: In the past I've always used /30's for PTP connection subnets out of old habit (i.e. Ethernet that won't take unnumbered) but now I'm considering switching to /31's in order to stretch my IPv4 space further. Has anyone else does this? Good? Bad? Based on the bit of testing I've done this shouldn't be a problem since it's only between routers. rfc3021 is over 9 years old, so should be no suprise that it works well. :-) Works well if supported. Vendor b (nee f) apparently dropped it off their roadmap. -- kris
Re: Using /31 for router links
Shouldn't be any issues...it's 2010 :) And, your IP allocation utilization will love you. tv - Original Message - From: Seth Mattinen se...@rollernet.us To: nanOG list nanog@nanog.org Sent: Friday, January 22, 2010 6:08 PM Subject: Using /31 for router links In the past I've always used /30's for PTP connection subnets out of old habit (i.e. Ethernet that won't take unnumbered) but now I'm considering switching to /31's in order to stretch my IPv4 space further. Has anyone else does this? Good? Bad? Based on the bit of testing I've done this shouldn't be a problem since it's only between routers. ~Seth
Re: Using /31 for router links
On 23/01/2010, at 1:31 PM, Jay Nugent wrote: Greetings, On Fri, 22 Jan 2010, Seth Mattinen wrote: In the past I've always used /30's for PTP connection subnets out of old habit (i.e. Ethernet that won't take unnumbered) but now I'm considering switching to /31's in order to stretch my IPv4 space further. Has anyone else does this? Good? Bad? Based on the bit of testing I've done this shouldn't be a problem since it's only between routers. Yes, this *IS* done *ALL* the time. P-t-P means that there are ONLY two devices on the wire - hence point to point. It ONLY uses two IP addresses (one on each end) and there is no reason or need to ARP on this wire. So no need for a broadcast or network addresses - it is just the two end points. ARP is still required on ethernet links, so that the MAC address can be discovered for use in the ethernet frame header. /31 does not change the behavior of ARP at all. -- Nathan Ward
Re: Using /31 for router links
Nathan Ward na...@daork.net wrote: ARP is still required on ethernet links, so that the MAC address can be = discovered for use in the ethernet frame header. /31 does not change the = behavior of ARP at all. soapbox That is why I hate Ethernet with a passion. Ethernet should be for LANs only; using Ethernet for WANs and PTP links is the vilest invention in the entire history of data networking in my opinion. My medium of choice for PTP links (WAN) is HDLC over a synchronous serial bit stream, with a V.35 or EIA-530 interface between the router and the modem/DSU. Over HDLC I then run either RFC 1490 routed mode or straight PPP (RFC 1662); in the past I used Cisco HDLC (0F 00 08 00 IP header follows...). My 4.3BSD router (or I should better say gateway as that's the proper 80s/90s term) then sees a PTP interface which has no netmask at all, hence the near and far end IP addresses don't have to have any numerical relationship between them at all. No netmask, no MAC addresses, no ARP, none of that crap, just a PTP IP link. /soapbox MS
RE: Using /31 for router links
ARP is still required on ethernet links, so that the MAC address can be discovered for use in the ethernet frame header. /31 does not change the behavior of ARP at all. -- Nathan Ward I often manually configure the MAC addresses in static fashion on point-to-points to eliminate the ARPing but that is nothing unique to a /31. It does eliminate the need for ARP, though.
Re: Using /31 for router links
On Sat, 23 Jan 2010 04:22:50 GMT msoko...@ivan.harhan.org (Michael Sokolov) wrote: Nathan Ward na...@daork.net wrote: ARP is still required on ethernet links, so that the MAC address can be = discovered for use in the ethernet frame header. /31 does not change the = behavior of ARP at all. soapbox That is why I hate Ethernet with a passion. Ethernet should be for LANs only; using Ethernet for WANs and PTP links is the vilest invention in the entire history of data networking in my opinion. My medium of choice for PTP links (WAN) is HDLC over a synchronous serial bit stream, with a V.35 or EIA-530 interface between the router and the modem/DSU. Over HDLC I then run either RFC 1490 routed mode or straight PPP (RFC 1662); in the past I used Cisco HDLC (0F 00 08 00 IP header follows...). My 4.3BSD router (or I should better say gateway as that's the proper 80s/90s term) then sees a PTP interface which has no netmask at all, hence the near and far end IP addresses don't have to have any numerical relationship between them at all. No netmask, no MAC addresses, no ARP, none of that crap, just a PTP IP link. /soapbox That's not a soapbox, that's a soap factory! What about NAT, ATM cell tax, unnecessary addressing fields in PTP protocols (including your beloved HDLC), SSAP, DSAP fields not being big enough in 802.2 necessitating SNAP, IPX directly over 802.3, AAL1 through AAL4, PPPoE dumbell MTUs and MSS hacks? Some of those are far worse sins in my opinion.