- Test - lists.nanog.org

Re: BBN outage
by Rob Liebschutz 15 Oct '96

15 Oct '96

> Most of the SF Bay Area portion of BBN Planet's network has been offline > since late last night. I've finally heard, via a friend who is a major > customer of theirs, that they are having power problems in Palo Alto. > > Now, I haven't heard a peep about this on any of the mailing lists I'm on... > and since knowing about BIG outages like this would make it easier for > me to answer my customer's questions... am I missing something? > > Is there a mailing list that I should be on? Or is this just a private > BBN issue that I'd need to call their ops center people about (people who > I'm sure are very very busy without small network operators calling them) > > -matthew kaufman > matthew(a)scruz.net > > I just spoke to their NOC and was told that a power switch that is supposed to be able to switch them between 3 different power utilities failed at 12:30 this morning (friday). They bought and installed 2 large diesel generators today and are hoping to be back up using the generators within 30 minutes. Rob

6 11

Re: BBN outage
by Dorian R. Kim 15 Oct '96

15 Oct '96

On Mon, 14 Oct 1996, Patrick J. Chicas wrote: > On Mon, 14 Oct 1996, Rob Gutierrez wrote: > > > 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. > > Their HUB is connected by one microwaved DS-3? This is not good. Granted, This DS3 is one of 4 private interconnect BBN has to MCI. I don't think there's anything wrong with it's current provisioning given that their other backbone DS3 have diverse routing.. Please correct me if I've missed something. -dorian rk, or 2) Was not available ... See my previous comments about rusty generators. > 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. Their HUB is connected by one microwaved DS-3? This is not good. Granted, most DS-3 microwave radios have hot standby radios, muldems ect. on each end but, they are always ripe for outage due to interference or path fade. I can see the microwave as a segment of a SONET OC-3 link, only if the other side is fiber. > 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. This is good.. How many of BBN's peer connection run over the SONET ring? Regards Patrick J. Chicas Email: pjc(a)unix.off-road.com URL: http://www.Off-Road.com -------------------------------- The Off-Road Center of The 'Net!

2 1

Re: Cisco's AIP vs HSSI
by salo＠msc.edu 15 Oct '96

15 Oct '96

> Date: Fri, 11 Oct 1996 15:09:51 -0400 (EDT) > From: Pritish Shah <pritish(a)iocenter.net> > To: nanog(a)merit.edu > Subject: Cisco's AIP vs HSSI > > Can anyone tell me which is better and why? Is there any resources out > there that will help me in that decision. The short answer: You will undoubtedly be much happier if you use the Cisco AIP card rather than a HSSI card with an external ATM DSU. A longer answer: HSSI ---- The HSSI card is a fairly dumb interface. It is actually an interface between the Cisco router and an external DSU. As I recall, it has a V.35 interface. I think its top speed is a bit more than a DS-3. Its original function was to connect a DS-3 (45 Mbps) DSU to a Cisco router. ---------- | | ------------ | Router |======| DS-3 DSU |----------> DS-3 leased line | | ------------ ---------- The first available method of connecting a Cisco router to an ATM network was to use an ATM DSU connected to a HSSI interface. The ATM DSU appeared to the Cisco router to be pretty much the same as a traditional DSU. All of the ATM functions, largely segmentation and reassembly, were performed in the ATM DSU. ---------- | | ------------ | Router |======| ATM DSU |----------> DS-3 leased line to ATM Network | | ------------ ---------- The HSSI card generally uses "frame relay" or NLPID encapsulation within the AAL frame. The more standard encapsulation today is the LLC/SNAP encapsulation. I believe that Digital Link built the first ATM DSU. This was a fairly early device, so it implemented AAL3/4, (rather than AAL5 which is the predominant AAL today). The Digital Link cannot use AAL5. The Cisco AIP card generally wants to use AAL5. We have been unsuccessful in connecting a Cisco with a HSSI and Digital Link to a Cisco with an AIP card. The details are another long discussion which I won't bother to type at this time. ADC Kentrox also builds an ATM DSU. It supports AAL5. However, the experience at the ATM NAPs indicate that this device has some performance limitations. Among other things, the DXI specification, (which specifies the interface between the router and the ATM DSU), says that only 16 reassembly buffers are required in the ATM DSU, (if memory serves me...). This limited reassembly space can cause problems. On the other hand, the Kentrox ATM DSU supports AAL5. I think it also interoperates with the Cisco AIP card, although I haven't seen it first hand. Other issues with a HSSI/ATM DSU solution include: o The DXI interface supports only PVCs. In a world which is moving towards SVCs, this is a rather serious limitation. o There are a whole host of new ATM features which the HSSI card/ADSU solution won't be able to support. o Check with Cisco about performance limitations in the HSSI card. o I don't believe that the HSSI card will support anything near OC-3c speeds. OC-3c ATM products are readily available in local area networks. Also note that most people run their wide-area ATM connection through a local ATM switch before connecting it to a router. I would be inclined to make the connection between the local ATM switch and the router OC-3c rather than DS-3 (for a bunch of reasons). Cisco AIP --------- I don't believe that the Cisco AIP card is the ultimate solution, but it is far superior to the HSSI interface. Important advantages include: o The AIP card supports SVCs. o The AIP card supports standard encapsulations, AAL5 and LLC/SNAP. (It might also support NLPID and AAL3/4; check with Cisco.) o The AIP supports a variety of interfaces. I think these include multi-mode OC-3c, single-mode OC-3c and DS-3. Verify with Cisco. o I think the performance of the AIP card may be better than the HSSI card, but I don't have any test results in front of me. Summary ------- Read my short answer and skip the rest. Disclaimer: Verify the facts with your vendor. I typed this in on the fly late at night; the general ideas are most likely correct, but some of the details might be a bit off. -tjs

1 0

Re: Cisco's AIP vs HSSI
by salo＠msc.edu 15 Oct '96

15 Oct '96

> Date: Fri, 11 Oct 1996 15:09:51 -0400 (EDT) > From: Pritish Shah <pritish(a)iocenter.net> > To: nanog(a)merit.edu > Subject: Cisco's AIP vs HSSI > > Can anyone tell me which is better and why? Is there any resources out > there that will help me in that decision. The short answer: You will undoubtedly be much happier if you use the Cisco AIP card rather than a HSSI card with an external ATM DSU. A longer answer: HSSI ---- The HSSI card is a fairly dumb interface. It is actually an interface between the Cisco router and an external DSU. As I recall, it has a V.35 interface. I think its top speed is a bit more than a DS-3. Its original function was to connect a DS-3 (45 Mbps) DSU to a Cisco router. ---------- | | ------------ | Router |======| DS-3 DSU |----------> DS-3 leased line | | ------------ ---------- The first available method of connecting a Cisco router to an ATM network was to use an ATM DSU connected to a HSSI interface. The ATM DSU appeared to the Cisco router to be pretty much the same as a traditional DSU. All of the ATM functions, largely segmentation and reassembly, were performed in the ATM DSU. ---------- | | ------------ | Router |======| ATM DSU |----------> DS-3 leased line to ATM Network | | ------------ ---------- The HSSI card generally uses "frame relay" or NLPID encapsulation within the AAL frame. The more standard encapsulation today is the LLC/SNAP encapsulation. I believe that Digital Link built the first ATM DSU. This was a fairly early device, so it implemented AAL3/4, (rather than AAL5 which is the predominant AAL today). The Digital Link cannot use AAL5. The Cisco AIP card generally wants to use AAL5. We have been unsuccessful in connecting a Cisco with a HSSI and Digital Link to a Cisco with an AIP card. The details are another long discussion which I won't bother to type at this time. ADC Kentrox also builds an ATM DSU. It supports AAL5. However, the experience at the ATM NAPs indicate that this device has some performance limitations. Among other things, the DXI specification, (which specifies the interface between the router and the ATM DSU), says that only 16 reassembly buffers are required in the ATM DSU, (if memory serves me...). This limited reassembly space can cause problems. On the other hand, the Kentrox ATM DSU supports AAL5. I think it also interoperates with the Cisco AIP card, although I haven't seen it first hand. Other issues with a HSSI/ATM DSU solution include: o The DXI interface supports only PVCs. In a world which is moving towards SVCs, this is a rather serious limitation. o There are a whole host of new ATM features which the HSSI card/ADSU solution won't be able to support. o Check with Cisco about performance limitations in the HSSI card. o I don't believe that the HSSI card will support anything near OC-3c speeds. OC-3c ATM products are readily available in local area networks. Also note that most people run their wide-area ATM connection through a local ATM switch before connecting it to a router. I would be inclined to make the connection between the local ATM switch and the router OC-3c rather than DS-3 (for a bunch of reasons). Cisco AIP --------- I don't believe that the Cisco AIP card is the ultimate solution, but it is far superior to the HSSI interface. Important advantages include: o The AIP card supports SVCs. o The AIP card supports standard encapsulations, AAL5 and LLC/SNAP. (It might also support NLPID and AAL3/4; check with Cisco.) o The AIP supports a variety of interfaces. I think these include multi-mode OC-3c, single-mode OC-3c and DS-3. Verify with Cisco. o I think the performance of the AIP card may be better than the HSSI card, but I don't have any test results in front of me. Summary ------- Read my short answer and skip the rest. Disclaimer: Verify the facts with your vendor. I typed this in on the fly late at night; the general ideas are most likely correct, but some of the details might be a bit off. -tjs

1 0

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