- Test - lists.nanog.org

Re: IPSEC and PAT
by Steven M. Bellovin 14 Sep '01

14 Sep '01

In message <LCEKLACNFGLMOPOGNBNMMEBBCEAA.tim(a)eng.bellsouth.net>, "Tim Irwin" wr ites: > >I looked at this a while back... I am dusting off the cobwebs of my mind, so >no flames please. I believe that the NATing device must modify the SPI >values. The sending device sends out an ESP packet with src addy of, say >192.168.1.2, to the NAT router. The router must look at the TCP port to >determine that it's IPSEC in order to figure out that it's a special case >and NAT it. It then must modify the SPI value (which is partially made up >of the src IP address) as it leaves because the NAT dst device will use the >info in the SPI value in the formulation of it's reply. > >If this is wrong, please correct me... I'm interested in knowing as well. That doesn't work -- the SPI is protected by ESP's authentication check (section 2 of RFC 2406) or by AH (section 2 of RFC 2402). --Steve Bellovin, http://www.research.att.com/~smb http://www.wilyhacker.com

1 0

Re: IPSEC and PAT
by Tony Rall 14 Sep '01

14 Sep '01

On Thursday, 2001/09/13 at 21:43 AST, "Steven M. Bellovin" <smb(a)research.att.com> wrote: > I repeat -- it doesn't do PAT. Some "routers" -- they're really no > such thing, of course; they're NAT boxes and/or bridges -- allow one > host behind them to speak IPsec. If a host emits a packet using ESP, > it's tagged as *the* IPsec user; return IPsec packets are routed to > that host. (Some of these boxes may use manual configuration instead > or in addition.) You can't have two IPsec hosts, because there's no > way to know which should receive incoming packets -- there's no > relationship between inbound and outbound SPIs. Actually you can have multiple IPSEC sessions hidden behind a NAT box with a single public IP address - we've found several vendors' "routers" that can work in this environment. I believe the key is that each tunnel must be to distinct remote IP addresses. All the NAT box has available to separate the traffic for the different tunnels (which use IP protocol 50) is the address of the other end of the tunnel, but that is all it needs. Of course, many users would like to have multiple tunnels to the same partner. I don't know how that is possible with current IPSEC technology. Tony Rall

4 3

Further considerations for the future
by George William Herbert 14 Sep '01

14 Sep '01

I posted on Wednesday regarding some long term network issues that I thought were raised by the NYC attacks. An additional one has occurred to me from some subsequent events. For organizations which are dependent on the net for day to day operations (despite the dot-com collapse, the number of these is steadily increasing) the acceptability of single homing is likely to have decreased significantly. I addressed this a bit with the significance of true separation in physical links that I brought up in the first mail. Beyond that, though, we have seen that whole facilities can be outright destroyed or fail due to cascading infrastructure and backups failures as in 25 Broadway (are your generators n+1 redundant...and do you stock spare water pumps for them?). This really means that except for truly extraordinary ISPs, no end customer can really be sure that an ISP is going to be able to maintain services through a major incident (or even a localized serious incident). Expect to see in the coming months many more end customers desiring to get truly dually homed, with redundant ISPs, redundant local loops in different directions, etc. We should be prepared to deal with the additional routing and provisioning complexity involved in handling additional customers doing BGP, dual homing, etc., both new customers coming for their second connection and existing customers as they reach out to a second ISP for the redundant connectivity. -george william herbert gherbert(a)retro.com

1 0

Emergency Alert System TEST - This is only a TEST
by Sean Donelan 14 Sep '01

14 Sep '01

FEMA will be conducting a TEST of the primary entry point system for the Emergency Alert System. The Emergency Alert System is the successor to the Emergency Broadcast System, and is frequently used for tornado warnings, hurricane warnings, flash floods warnings and so forth. It is also designed, but never in history used, for Presidential messages to the nation. The FCC Compliance Bureau has suspended all other testing of the Emergency Alert System for 30 days. However, some automated broadcast stations may send a test message without operator intervention. If the EAS is activated in your area listen carefully if it is a test or a real activation. The EAS has been accidently activated in the past, so if it is activated, avoid undue panic and listen carefully. Severe weather alerts using the EAS are continuing. Most broadcast AM/FM, television and cable systems participate in the Emergency Alert System. However, participation is voluntary. Individual stations may or may not be participate. Non-participating stations will go off the air during a national EAS activation. Once again, this is a TEST. This is only a TEST. > GREETINGS: > > THE PRIMARY ENTRY POINT SYSTEM WILL BE TESTED IN THE NEXT FEW DAYS. THE > TESTS WILL BE CLOSED CIRCUIT AT BEST. THERE MAY BE SOME AUDIO PASSED AS > PART OF THE TEST. IN ONE CONFIGURATION THE PEP CONFERENCE IS INVOKED, BUT > ENDECS ARE NOT ADDRESSED OR POLLED. > > DAVE PIPER AND I HAVE GONE OVER THE TEST PLAN TO CONFIRM THAT IT IS WITHIN > THE PEP SPECIFICATIONS. REST ASSURED THAT THE PEP SYSTEM WILL NOT BE > ACTIVATED WITHOUT PROPER AUTHORIZATION AND APPROVAL. > > I WILL TRY TO MAKE THE RESULTS OF THE TEST AVAILABLE VIA THIS SITE, TIME AND > WORKLOAD PERMITTING. > > I HAVE HAD A LOT OF E-MAILS ON THE TESTING, THANK YOU FOR YOUR COMMENTS AND > INPUT. CIRCUMSTANCES ARE SUCH THAT THE TESTING MUST BE ACCOMPLISHED AT THIS > TIME. > > PUTPRUSH SENDS > > TIMOTHY L. PUTPRUSH, FEMA PEP PROJECT MANAGER

1 0

Re: IPSEC and PAT
by Steven M. Bellovin 14 Sep '01

14 Sep '01

In message <912A91BC69F4D3119D1B009027D0D40C01BB45A7(a)exchange1.secure.insweb.co m>, Vandy Hamidi writes: >It is working now. I've done it with Linksys and Netopia DSL routers. >Software client on the laptop that DOES tunnel mode ESP. No AH and running >through a PAT and it works flawlessly. I just want to know how it works, >I've already determined that it does. >The point where my logic fails is where PAT relies on modifying the TCP/UDP >port numbers, an ESP packet has a standard IP header with an additional >protocol 50 ESP header. Since there is no ports to change to create a table >to keep track of which packet came from which internal client, what is used >to keep track. >Someone said something about the UDP encapsulation, but what about the >NETOPIA which doesn't do that? I repeat -- it doesn't do PAT. Some "routers" -- they're really no such thing, of course; they're NAT boxes and/or bridges -- allow one host behind them to speak IPsec. If a host emits a packet using ESP, it's tagged as *the* IPsec user; return IPsec packets are routed to that host. (Some of these boxes may use manual configuration instead or in addition.) You can't have two IPsec hosts, because there's no way to know which should receive incoming packets -- there's no relationship between inbound and outbound SPIs. As for the UDP encapsulation -- yes, the IETF's IPsec working group is moving in that direction. But it's not standardized yet, and there may be patent issues to sort through. --Steve Bellovin, http://www.research.att.com/~smb http://www.wilyhacker.com

2 1

Latest from Telehouse...
by Jeffrey Meltzer 14 Sep '01

14 Sep '01

The water pump on their generator is broken. They don't expect to have it fixed until tomorrow night. They are however, expecting a loaner Generator from Con Ed they they hope to have in their possession in the next few hours and setup by morning...

1 0

Assistance Available
by rbeach＠nycap.rr.com 14 Sep '01

14 Sep '01

I'm teaching at a small school here in upstate NY......and I have a bunch of eager students and would like to organize some help. We can offer bodies to run wire, connect, install equipment.....contact me off-line and we can plot and scheme. Richard Beach rbeach(a)nycap.rr.com 518.265.7434

1 0

Re: telehouse - 25 broadway
by Marshall Eubanks 14 Sep '01

14 Sep '01

> >>Power was restored to our equipment at 02:28 (UK) > >...and lost again at 02:45 (UK). > >Perhaps I'll keep quiet next time. Simon; I do not seem to have any multicast connectivity to the BBC since the IETF. Did you take mbgp down ? We have not had multicast connectivity to Dante since 24 bway went down, and I thought that is why we lost the BBC, but this predates this by a good margin. We found that only multicast functioned well during the tremendous load experienced on Tuesday. Do you have any multicast news feeds up ? I was deluged with requests for news after the general melt down on Tuesday, and would be glad to publicize any multicast news feeds sourced from the BBC. I do not think that this is going away in the near future. Marshall > >Simon >-- >Simon Lockhart | Tel: +44 (0)1737 839676 >Internet Engineering Manager | Fax: +44 (0)1737 839516 >BBC Internet Services | Email: Simon.Lockhart(a)bbc.co.uk >Kingswood Warren,Tadworth,Surrey,UK | URL: http://support.bbc.co.uk/ > > > Marshall Eubanks tme(a)21rst-century.com

1 0

(fwd) Re: telehouse - 25 broadway
by Simon Lockhart 14 Sep '01

14 Sep '01

------- Forwarded Message Date: Thu, 13 Sep 2001 22:06:14 -0400 From: Steven Nash <snash(a)lightning.net> Subject: Re: telehouse - 25 broadway Simon, Please post to list, never emailed for posting access :-) The power to at least parts of the 5th floor was returned until about 10 minutes ago (9:47 EDT). The power as of now is as it was, completely gone as we await for the generator fix. No ETR I know of at this time. At 02:29 AM 9/14/2001 +0100, you wrote: > >well their generator overheated. no power at the facility > >Status as at 02:25 (UK) / 21:25 (New York): > >They have a Generator Tech on site at the moment, and he is working on >fixing the generator. They have also spoken to Con Ed about borrowing a >generator to get them going. Someone from Con Ed is heading down to see >what they can do. Telehouse seem pretty hopeful that they will get power >back in a few hours. > >Simon >-- >Simon Lockhart | Tel: +44 (0)1737 839676 >Internet Engineering Manager | Fax: +44 (0)1737 839516 >BBC Internet Services | Email: Simon.Lockhart(a)bbc.co.uk >Kingswood Warren,Tadworth,Surrey,UK | URL: http://support.bbc.co.uk/ - ---- Steven Nash Director of Network Engineering Lightning Internet Services, LLC http://www.lightning.net o:516.248.8400 f:516.248.8897 c:516.779.3007 e:888.935.2698 "remember, hijackers are ticketed passengers too." ------- End of Forwarded Message -- Simon Lockhart | Tel: +44 (0)1737 839676 Internet Engineering Manager | Fax: +44 (0)1737 839516 BBC Internet Services | Email: Simon.Lockhart(a)bbc.co.uk Kingswood Warren,Tadworth,Surrey,UK | URL: http://support.bbc.co.uk/

1 0

RE: IPSEC and PAT
by Vandy Hamidi 14 Sep '01

14 Sep '01

I see now, thanks! My logic is complete. thanks, -=Vandy=- -----Original Message----- From: Steven M. Bellovin [mailto:smb@research.att.com] Sent: Thursday, September 13, 2001 6:44 PM To: Vandy Hamidi Cc: nanog(a)merit.edu Subject: Re: IPSEC and PAT In message <912A91BC69F4D3119D1B009027D0D40C01BB45A7(a)exchange1.secure.insweb.co m>, Vandy Hamidi writes: >It is working now. I've done it with Linksys and Netopia DSL routers. >Software client on the laptop that DOES tunnel mode ESP. No AH and running >through a PAT and it works flawlessly. I just want to know how it works, >I've already determined that it does. >The point where my logic fails is where PAT relies on modifying the TCP/UDP >port numbers, an ESP packet has a standard IP header with an additional >protocol 50 ESP header. Since there is no ports to change to create a table >to keep track of which packet came from which internal client, what is used >to keep track. >Someone said something about the UDP encapsulation, but what about the >NETOPIA which doesn't do that? I repeat -- it doesn't do PAT. Some "routers" -- they're really no such thing, of course; they're NAT boxes and/or bridges -- allow one host behind them to speak IPsec. If a host emits a packet using ESP, it's tagged as *the* IPsec user; return IPsec packets are routed to that host. (Some of these boxes may use manual configuration instead or in addition.) You can't have two IPsec hosts, because there's no way to know which should receive incoming packets -- there's no relationship between inbound and outbound SPIs. As for the UDP encapsulation -- yes, the IETF's IPsec working group is moving in that direction. But it's not standardized yet, and there may be patent issues to sort through. --Steve Bellovin, http://www.research.att.com/~smb http://www.wilyhacker.com

1 0