- Test - lists.nanog.org

Re: Building Evacuation [Re: BBN outage]
by Sean Donelan 15 Oct '96

15 Oct '96

>You have to remember that it's harder to keep people in a given >building than it is to safeguard the equipment there. You always >have to be ready to send your NOC operators packing with a cell >phone and a laptop for up to 48 hours, in case the building has >to be evacuated for one of a dozen reasons. You should have >a backup call center with an ACD set up in some other city or >a contract with someone for a backup service. You realize you have just cursed someone's NOC to this fate. If my next post comes from Monterey, California, you'll know we had to evacuate to our backup NOC. Once upon a time, when Merit ran the Internet, as their last best hope they had a plan for the NOC operators to grab a laptop as they ran from the NOC. One of the harder things to plan is how you are going to get back inside your possibly non-functioning, or damaged network from the outside without leaving too many backdoors for the evil ones. Its a embarassing when your SecureID servers crashes inside a building you can't physically enter, and none of your network equipment will let you log in. Its even more embarassing when a cracker uses your 'secret' backdoor and takes over your network. -- Sean Donelan, Data Research Associates, Inc, St. Louis, MO Affiliation given for identification not representation

1 0

Re: generators, etc....
by Paul Ferguson 15 Oct '96

15 Oct '96

At 08:11 PM 10/12/96 -0500, Sean Donelan wrote: > >Stuff happens. No one can predict or plan for everything. But you can >have procedures in place to combat the biggest problem in diasasters, poor >communication. Partial knowledge about the real situation happens when >the real information isn't made available. Techies sometimes get too >wrapped up in the hardware, redudant fiber, generators, bomb-resistant >shelters; and forget about keeping people informed. > Here's what the latest issue of RISKS Digest [18.52] had to say. - paul [snip] ---------------------------------------------------------------------- Date: Sat, 12 Oct 1996 13:48:41 -0700 From: "Peter G. Neumann" <Neumann(a)csl.sri.com> Subject: Rats take down Stanford power and Silicon Valley Internet service Two rats crawled through an underground cable conduit into a cabinet of power switching gear adjacent to the Stanford University cogeneration plant, and caused an explosion that cut off power to the Stanford area beginning around 7:30pm on Thursday evening, 10 Oct 1996, and continuing until 3:30pm Friday afternoon. The BBN Planet hub (Internet Point of Presence, or PoP) at the Stanford University Data Center remained in operation for a few hours on standby battery power, but then gave out around 9pm Thursday; it came back up around 4:30pm, an hour after Stanford restored power. To name just a few, Bay-Area BARNet users at Stanford, U.C. Berkeley, Apple, Sun, Hewlett-Packard, Lawrence Livermore (partially), and SRI were cut off from the Internet. The *Los Angeles Times* and *San Francisco Chronicle* on-line sites were also off the air. Because I had no Internet access yesterday, I held up RISKS-18.52 -- thus enabling me to include this item adding to our RISKS archives collection of rodent-induced outages. (Long-time readers recall that SRI alone has contributed four fresh-fried squirrels resulting in power outages.) [Sources: On-line messages and a front-page *San Francisco Chronicle*, 12 Oct 1996 item] Evidently, the horse is out of the BARNet, and the rats found the weak lynx. They sure put a ro-dent in the day for many BayAreans. Perhaps your mouse will click on a tale of its own. At any rate, this is just one more saga in the weak-link-in-the-infrastructure department. But I'm surprised that power-system technology has not found a way to develop rodent-tolerant circuits. [With SysAdmins and others pacing the halls at SRI waiting for whatever, Doug Moran remarked that keeping around a few fresh-frozen electrofried rodents is allegedly standard practice for purveyors of power; it is then very easy to have a fallback alibi when no other cause can be found.] ------------------------------ [snip]

2 1

Re: cable modems [Re: Intra/Inter - was Inet-II]
by Sanjay Dani 15 Oct '96

15 Oct '96

> The power of the KPCB media hype machine. There is an interesting > and rather lengthy article in this month's Upside finally questioning > the long term viability of @Home from a financial point of view. > The issue turns on the lack of investment on the part of TCI to make > the necessary cable plant upgrades. No upgrades -- no one @Home. > > The hype machine always unravels, given time. Yeah, only after consuming a few class B's and a class A reservation through special channels ;-) Seriously, KPCB won't let this one fail that easily. Other KPCB (successful) startups are known to be promoting the larger cause of the "keiretsu" by helping this one out in their own ways. Sorry, off topic but couldn't resist.

1 0

Re: Cisco's AIP vs HSSI
by Dorian R. Kim 15 Oct '96

15 Oct '96

On Mon, 14 Oct 1996, Kent W. England wrote: > At 03:09 PM 11-10-96 -0400, Pritish Shah wrote: > > > >Can anyone tell me which is better and why? Is there any resources out > >there that will help me in that decision. I'd actually suggest looking at http://www.cisco.com/ > Basically, the cisco AIP comes with an ATM interface. If you buy > the HSSI then you need an ATM CSU/DSU to connect to the ATM network. > > My personal opinion is that the cisco AIP is simpler to use. However, this is not necessarily the case when your objective doing none ATM things. -dorian

1 0

generators, etc....
by mo＠UU.NET 15 Oct '96

15 Oct '96

Even when you have generators, there are still a myriad of things which can and do go wrong no matter how hard you try, even with trebly redundant backups, etc. etc. A fuel supplier accidently puts half a load of jet fuel instead of #2 diesel in your storage tank, which you won't find until it reloads the run tank from the day tank. Floods happen and the water rises taller than the 3 story snorkle on the diesels. Generators takes a direct lightning strike and fries house DC power (even inside shielded enclosures). A fiber transmission system goes crazy when the control system is zapped by the lightning strike on the generator. A 200mph hurricane gust rips a microwave system off the roof (tower and all) and throws it down on the generators, crushing the exhaust system and the diesels strangle. (No, I'm' not imagining these.) We all try very, very hard to make things reliable, but the world isn't perfect, nor are any of us. BBN will probably find some things to improve and change the odds next time. Then it will be someone else's turn to catch the javelins. I suggest that neither gleefull hand-rubbing nor "obvious" pronouncements based on partial knowledge about the real situation will seem quite so appropriate when it's *your* turn to be downfield from the launcher. Peace, -mo

8 10

cable modems [Re: Intra/Inter - was Inet-II]
by Kent W. England 15 Oct '96

15 Oct '96

At 12:18 PM 11-10-96 -0700, maillists wrote: >The same article quoted someone at EFF that there is no need >for a new Internet as the current one can be made faster by using >cable modems. Hmmm.. > >Sanjay. > > > > The power of the KPCB media hype machine. There is an interesting and rather lengthy article in this month's Upside finally questioning the long term viability of @Home from a financial point of view. The issue turns on the lack of investment on the part of TCI to make the necessary cable plant upgrades. No upgrades -- no one @Home. The hype machine always unravels, given time. --Kent

1 0

Re: Creating exchanges
by Barry Raveendran Greene 14 Oct '96

14 Oct '96

Hello All, This topic of Guam as an IX has come up many times in the past. Frankly, if it made since, it would have already happened. The #1 factor for an IX location in the Asia Pacific is the business case. Geographic topology does play a factor, but 1/2 circuit lease line prices are the major determinant if people will come to an IX. Guam's problem is that people have to pay the full circuit cost to get to Guam. These circuit costs are roughly 10X the cost of circuits in the US. In addition, pricing of the circuits are NOT distance sensitive. So for most of the region, the full circuit costs to Guam will be more expensive than a circuit to the US. One engineer/colleague working for a large US Internet backbone did work through the business case of placing a major hub in Guam for his Internet, FR, and X.25 traffic. He had the advantage of knowing what 'military' resources he can draw from (i.e. Typhoon hardened FM facilities) case he use to work for the Air Force in the region. The results was that the business case did not make since. Using Japan, HK, Singapore, and Australia as redundant hubs made better since. And that is what happened when they upgraded their backbone. Robert Mathews-ICICX wrote: >> From an infrastructure perspective, though, Japan looks hard to beat. >> The pipes to Singapore/Jakarta/Australia, Guam/Hawaii/L.A.-ish, and >> somewhere in Oregon (?) all meet there. Ignoring regulations, tariff >> issues, etc., of course. Japan and Singapore does look good because of the cable interconnects and the satellite foot prints. AIH/Abone (Japan - http://www.aih.net) and STIX/SIB (Singapore - http://www.stix.net) are two successes of international IXs - through both are commercial IXs - not neutral. Yet, each of them are working. Personally, after working this issue for 3 1/2 years, international neutral IXs will not work in Asia Pacific. National IXs are fantastic and work wonders. Commercial IXs are there to compete with the big US backbones in the region. Barry -- Barry Raveendran Greene | || || | Senior Consulting Engineering | || || | Singapore | |||| |||| | tel: +65 738-5535 ext 235 | ..:||||||:..:||||||:.. | e-mail: bgreene(a)cisco.com | c i s c o S y s t e m s |

3 2

Re: Internet: Getting more money ...??
by Dana Hudes 14 Oct '96

14 Oct '96

Point may have been obscured. You add more facts (thanks!) the question is what the $10 million represented -- did it include the money IBM spent but did not charge the NSF? As for the NSS's being still in place but disconnected, I thought the lease was up a year ago. The NSS's certainly are not cost effective. One of the problems that they faced in the versions subsequent to those you had at ANS was running out of backplane bandwidth for WAN-LAN (and LAN-LAN) transfers. Another of course was their proprietary nature. The price tag doesn't surprise me, Rs/6ks are expensive. But the sampling feature would be helpful in the SYN attack situation. Dana

2 1

Digest Possible??
by Andrew Smith 14 Oct '96

14 Oct '96

This question has probably been asked many times over, but I haven't been able to find it or an answer. Is there, and if there isn't, is getting it setup possible, a nanog-digest mailing list. --------------------------------------------------------------------------- Andrew Smith ** awsmith(a)neosoft.com ** Network Engineer ** 1-888-NEOSOFT ** "Opportunities multiply as they are seized" - Sun Tzu ** ** http://www.neosoft.com/neosoft/staff/andrew ** ---------------------------------------------------------------------------

1 0

Re: BBN outage
by rmg＠ranma.com 14 Oct '96

14 Oct '96

From: Nathan Stratton <nathan(a)netrail.net> > > So BBN says they are now going to install a generator, my question is why > was a generator not installed? They saved 20 - 100K on a generator, but I > bet they will lose more then that when users leave. Forsythe Hall, where the BBN pop is located, has a standby generator, abet not hooked up, to provide power when the Stanford CoGenerator facility goes off, like it did. The bldg UPS has plenty of power to keep the facilities running until the standby generator can be hooked up (BBN shares physical facilities with Stanford's IBM mainframe and the Stamford DMS-100 telephone switch). When I used to work there, I saw that generator all the time. It collected a lot of dust. I was told that it was tested out from time to time :) I wonder why it either 1) Did not work, or 2) Was not available ... Across the street, at the hospital, there is another generator set up on a trailer. When the power was out, that one was not running. I would have thought that could have been disconnected and brought over to Forsythe Hall to provide power for when the UPS's died off. Telco wise, BBN does have a nice dirverse system. Their MCI T-3 is on a microwave that goes to Palo Alto, where it drops off to MCI's bay area fiber ring between San Jose to San Francisco. Some of their backbone T-3's come into a Pac Bell OC-48, which has an entirely seperate protection path into Stanford itself (Stanford alltogether has two Pac Bell OC-48's, with a third one planned!). Another fiber path belongs to MFS for more dirversity. So on that level, they have their act together. r.

1 0