October 2003 - Test - lists.nanog.org

Re: Finding clue at comcast.net
by Sirius F. Crackhoe 09 Oct '03

09 Oct '03

-----Original Message----- From: Sirius F. Crackhoe [mailto:sirius@crackhoe.com] Sent: Thursday, October 09, 2003 12:37 PM To: 'Howard C. Berkowitz' Comcast's Technical and Customer Support is outsourced to EDS and is based in the EDS call center in Hallifax. I believe they were also taking calls in the Winchester, KY call center, which was an old MCI/WorldCom Outsourced Call Center that they sold to EDS a few years back. They are setup the same as MSN, WebTV and RoadRunner. I used to do implementations for WorldCom's outsourced call centers and call tell you for sure, they have no access to the NOC or their staff. They are merely paid phone operators. :) Sirius -----Original Message----- From: owner-nanog(a)merit.edu [mailto:owner-nanog@merit.edu] On Behalf Of Howard C. Berkowitz Sent: Thursday, October 09, 2003 11:21 AM To: nanog(a)merit.org At 9:29 AM -0500 10/9/03, Austad, Jay wrote: >Comcast's phone support department is the *worst*, WORST, I've ever >dealt with. I think they are outsourced, they have to go by a script, >and many of them probably hardly know what a computer even is. Once I >called because of a problem on their network, and I told the person on >the phone that there was a problem on their network, and I pinned it >down to a couple of routers where the problem may be, and she >responded, very sternly, "Sir, WE DON'T HAVE ANY ROUTERS" Same thing here. Last night, I was told that no escalation personnel were available. On the couple of occasions where I got escalation, I once had an informal conversation with a 3rd level. Their phone center is in Halifax, NS -- didn't find out if it is outsourced or not. While the person with whom I spoke was reasonably clueful, he told me that customer support had no interactive communication with network operations -- at best, they could send an email about a routing, SMTP, etc. problem and hope somebody would respond. At the time, I was paying for their "Pro" service, intermediate between regular residential and full business. My contact said that while that was supposed to get better customer support, an early plan to route it to business Comcast failed, and there really was NO separate Pro support organization. I dropped the Pro service after I learned that residential service no longer insisted you remove any local routers and firewalls before deigning to troubleshoot. They still ask you to do that, but repeated NO responses can get them to proceed. A few NANOGs back (Atlanta), I did a presentation on customer satisfaction, which, frankly, was in many respects a case study of how I'd reform customer support at my then ISP/DSL, cais.net. If NANOG ever did formal documents, I'd like to see a guideline on how to run customer support. > >In any case, if you manage to get the call escalated a couple of times >(after lying about rebooting your computer 47 times), You forgot reinstalling Windows. On a Mac. >you'll get someone >good. Also, there are some good people who read this list. But >calling their phone support to get anything useful is like trying to >squeeze blood from a rock. > >-jay > >> -----Original Message----- >> From: Howard C. Berkowitz [mailto:hcb@gettcomm.com] >> Sent: Wednesday, October 08, 2003 7:36 PM >> To: nanog(a)merit.org >> Subject: Finding clue at comcast.net >> >> >> >> I'm rapidly beginning to believe this is equivalent to finding the >> pot of gold at the end of the rainbow. When my broadband alternative >> is Verizon and it's looking better, this is scary. >> >> Sometime today, their SMTP server started bouncing messages with >> more than 3 addressees. When I called customer support, I was told >> "we only handle troubleshooting, not mail service." The operator >> "guessed" they might be doing software updates on the mail service, >> had no information, and said there was no person to which it could be >> escalated. >> >> She insisted that I call my local cable office to find out when the >> "server repair" would be completed, because "they schedule all >> repairs." >> >> Is this a bad dream? >>

1 0

Customer support tutorial
by Howard C. Berkowitz 09 Oct '03

09 Oct '03

I was about to answer a request for my presentation about customer support, but I had a cat-on-keyboard exploit and I don't know who asked. In any event, the Atlanta presentation is at http://www.nanog.org/mtg-0102/cust.html I did it as a two-part with an intermediate BGP tutorial. In one of the two, there is an RFC 2270 slide. Please, please ignore it -- my brain went out to lunch on that one! Actually, this raises the interesting point -- is there an interest in updating and having running commentary on older presentations, keeping some content up to date? Is this something the NANOG site could support?

1 0

Why can't all my routers do FOO ?
by George M Jones 09 Oct '03

09 Oct '03

I'm trying to finish off the operational security requirements draft (http://www.ietf.org/internet-drafts/draft-jones-opsec-01.txt) which is a collection of operational security requirements for routers and other network infrastructure. The last major change that needs to happen is splitting out the Best Current Practice (BCP) info from the other items. This is where I'd like some feedback. If you're so motivated take 5 minutes to brainstorm two lists. The first is "*security features* I use everywhere now" (logging, aaa, filters...). The second is everything else: "I can't believe no vendor does FOO...", "Vendor A does BAR, life/security would be so much better if all vendors did it" etc. I'll take your lists and try to correctly align them with the drafts. If you're more motivated, you can see exactly which features have migrated from the BCP draft (opsec-01a-toc.txt) to the "info" draft (opsec-info-00.txt) by looking at the table of contents of the work-in-progress drafts @ http://www.port111.com/opsec/ Replies can come to the list (preferred to avoid duplication, allow discussion) or to me directly. Thanks, ---George Jones

1 0

News coverage, Verisign etc.
by Howard C. Berkowitz 09 Oct '03

09 Oct '03

I have gotten a reasoned response from the technology editor of the Washington Post, and we are discussing things. While I wouldn't have done it that way, he had a rational explanation of why the story was written the way it was, and definitely indicating there will be continuing coverage of the issue. He believes there's always room for improving coverage.

13 19

NANOG 29 hotels
by Simon Lockhart 09 Oct '03

09 Oct '03

Gah, the hotels seem to have suddenly all got booked up. Can anyone recommend a hotel close to NANOG29, sub $200/night room rate, with internet access in the room, with availability 18-22 Oct? I can't find any on the list on www.nanog.org with availability (or are > $200/night). Simon -- Simon Lockhart | Tel: +44 (0)1628 407720 (x37720) | Si fractum Technology Manager | Fax: +44 (0)1628 407701 (x37701) | non sit, noli BBC Internet Operations | Email: Simon.Lockhart(a)bbc.co.uk | id reficere BBC Technology, Maiden House, Vanwall Road, Maidenhead. SL6 4UB. UK

5 5

Transit and Paid Peering Exchanges
by Daniel Golding 09 Oct '03

09 Oct '03

Doing some research.... Anyone have a list of Transit and Paid Peering exchange fabrics? I am interested in both US and EU locations, particularly in interesting sites like 111 8th Ave (NYC), Telehouse North (London) and other major telco hotel type facilities. I'll summarize for the list and repost. Thanks, -- Daniel Golding

2 4

History of unannounced changes to the DNS core
by Sean Donelan 09 Oct '03

09 Oct '03

Try to match the quotes up with the actions. Postel gave no prior notification that the test would take place, a fact that drew criticism. "It's caused a good deal of uncertainty and perceived instability in the system," said Chris Clough, a spokesman for Network Solutions, which operates under a contract with the National Science Foundation. "It's a concern about who can authorize changes over the infrastructure and traffic patterns of the Internet." - Associated Press, February 4 1998 "We're very concerned," Becky Burr, administrator with the Commerce Department, said Tuesday. "This was undertaken without consultation with the United States government." People trying to visit the popular Internic directory which checks the availability of a new Web address are unexpectedly being swept automatically instead to the home page for Network Solutions Inc., which offers to register Internet addresses with the com, net or org suffixes for $119. - Associated Press, March 24 1999 NSI, naturally, has a different point of view. "The registry information is our proprietary information." says NSI spokesman Chris Clough. "We've been providing it free to the community, but under the contract, all of the intellectual property gathered through the InterNIC process is our proprietary information." - ABC News, March 7 1999

2 1

Contacts for CHINANET-BJ?
by Tom (UnitedLayer) 09 Oct '03

09 Oct '03

Anyone got a clueful contact over there? Getting 100Mbps or so of dos from over there and I'd rather not just blackhole the /16 --- Tom Sparks

1 1

i'm missing my copy of "why a wildcard MX won't help sitefinder"
by Paul Vixie 09 Oct '03

09 Oct '03

at the meeting yesterday, verisign said they were considering the benefits of a wildcard MX RR whose target was a nonexistent name, as a way to keep smtp traffic from having to come to sitefinder for rejection. i recall a very good posting on this topic and i think it was on nanog but i can't find it now. can someone privately send it to me if you've got it? -- Paul Vixie

1 0

[nanog@Overkill.EnterZone.Net: Extensions to RFC1998 - WAS: Re: DoS Attacks]
by Haesu 09 Oct '03

09 Oct '03

Forwarding to NANOG on behalf of Mr. Fraizer. Please don't shoot the messenger for any arguable/discussions. -hc -- Haesu C. TowardEX Technologies, Inc. Consulting, colocation, web hosting, network design and implementation http://www.towardex.com | haesu(a)towardex.com Cell: (978)394-2867 | Office: (978)263-3399 Ext. 170 Fax: (978)263-0033 | POC: HAESU-ARIN ----- Forwarded message from John Fraizer <nanog(a)Overkill.EnterZone.Net> ----- X-Original-To: haesu(a)mx01.bos.ma.towardex.com Delivered-To: haesu(a)mx01.bos.ma.towardex.com Date: Wed, 8 Oct 2003 21:58:26 -0400 (EDT) From: John Fraizer <nanog(a)Overkill.EnterZone.Net> To: Haesu <haesu(a)towardex.com> Cc: nanog(a)merit.edu, a_spicer(a)bellsouth.net Subject: Extensions to RFC1998 - WAS: Re: DoS Attacks In-Reply-To: <20031009004646.GB88086(a)scylla.towardex.com> X-Spam-Status: No, hits=-2.0 required=5.0 tests=EMAIL_ATTRIBUTION,IN_REP_TO,QUOTED_EMAIL_TEXT, REPLY_WITH_QUOTES,USER_AGENT_PINE version=2.55 X-Spam-Level: X-Spam-Checker-Version: SpamAssassin 2.55 (1.174.2.19-2003-05-19-exp) On Wed, 8 Oct 2003, Haesu wrote: > Hmmmm? What did I miss? When did RFC1998 get updated to include Null > community? Feel free to let me know if they updated RFC on that > lately.. b/c I havent checked it in a while. > > As far as I know, my upstreams are fully RFC1998 compliant and I use them well. > > -hc > Note: please echo this to the list. I don't have post access. Ahem... Sue...Ahem... The RFC itself hasn't been updated to include a Null community but if you think about it, providing a NULL community is fully within the concept of allowing customers to influence routing policy with the use of community strings. For example: ! router bgp 65534 neighbor a.a.a.a remote-as 65530 neighbor a.a.a.a description Customer AS65530 neighbor a.a.a.a prefix-list AS-65530 in neighbor a.a.a.a route-map CUSTOMERS in ! ip prefix-list AS-65530 seq 5 permit x.x.x.x/y le 32 ! ip community-list standard POISON permit 65534:666 ! route-map CUSTOMERS permit 10 match community POISON set local-preference 500 set ip next-hop [ip address of your sink-hole] ! Of course, the rest of the route-map CUSTOMERS is going to need to do some sanity checks on the announcements you accept from the customers OTHER than blackhole requests. In our case, we pass them through a prefix-list match that includes: ip prefix-list CUSTOMERS seq 10 deny 0.0.0.0/0 ge 25 As you can see, we're doing a prefix-list check against the announcements that the customer sends us in the neighbor statement. Each customer gets their own prefix-list that covers the networks that we have LOA to accept from that customer. (Keeps boneheads from blackholing OTHER people!) The first stanza in the CUSTOMERS route-map checks for the POISON community. Any prefix that the customer sends us that includes this community will be routed to our sink-hole. The rest of the stanzas in the CUSTOMERS route-map look for other communities from the customer. One stanza looks to see if the customer is requesting us to pass their announcements of our address space on as de-aggregated announcements. If we don't see that community, they're aggregated. Other stanzas in the route-map are pretty cut and dry RFC1998. Our customers can do the following: Community Action --------------------------------------------- 13944:0 Don't announce to any peer 13944:1 Don't announce to PEERS 13944:2 Don't announce to TRANSIT 13944:3 Don't announce to CUSTOMERS 13944:20 Announce specific from EnterZone aggregate for customers who are running on our IPs. 13944:90 Set preference to 90 13944:100 Set preference to 100 13944:110 Set preference to 110 13944:120 Set preference to 120 13944:666 Poison a Route 13944:NNN0 don't announce to Peer NNN 13944:NNN1 prepend once towards Peer NNN 13944:NNN2 prepend twice towards Peer NNN 13944:NNN3 prepend thrice towards Peer NNN Any time I do any consulting on another network, I recommend that they at MINIMUM implement the Poisoned Route ability. It is not terribly difficult to do as you can see above. -- John Fraizer EnterZone, Inc (13944+$|13944+_14813+$|13944+_17266+$) PGP Key = 6C5903C4 Fingerprint = 2AA6 6614 1B5E EDD2 38AD C417 3E61 F975 6C59 03C4 ----- End forwarded message -----

1 0