April 2001 - Test - lists.nanog.org

Carrier reliability and diversity
by Sean Donelan 05 Apr '01

05 Apr '01

The air line reservation system Galileo had significant problems earlier. Excerpt from Reuters: "The outage began about 8:30 p.m. Mountain Daylight Time on Wednesday and lasted until 10:55 a.m. MDT on Thursday, said Beth Tennis, vice present of enterprise networks for Quantitude, Galileo's Denver-based networking subsidiary. It affected three of the four T-1 high-speed data transmission lines supplied to Galileo by AT&T Corp., said Beth Tennis" Although I've found AT&T has the highest reliability of all the carriers I've used. Redundant links from the same carrier frequently have a common vulnerability, the carrier. For things which absolutely can't go down, I assume even AT&T will have occasional problems. Personally I used AT&T Worldnet and Earthlink, alternating depending on who had a problem. That's high praise for me.

1 0

Re: Major Qwest ATM outage.
by Sean Donelan 05 Apr '01

05 Apr '01

On Thu, 05 April 2001, "Kevin Oberman" wrote: > As was reported when Qwest did the software upgrade to all of their > Lucent (Ascend) 550s, the Qwest IP net does not use ATM. It is > entirely POS. As a result, the Qwest IP side was not impacted. > > This does look like the sort of like the cascading PNNI routing > problems that the big upgrade last week was supposed to have fixed. I > think the Lucent folk still have problems to deal with. Folks have reported that qwest said the cause for the failure was an OC48 going down, which led to instabilities within their network.

1 0

Re: Major Qwest ATM outage.
by Sean Donelan 05 Apr '01

05 Apr '01

On Wed, 04 April 2001, Scott Call wrote: > For any of us unlucky enough to have Qwest ATM service, they are > having a major outage. Started exactly at midnight eastern, and the MAST > person I spoke with made a reference to lucent conference calls, so I'm > betting on a bungled software upgrade or something similar. > > Qwest Master ticket # 219247. It doesn't seem to be affecting Qwest's IP network. Qwest's IP network status page reports 100% availability. http://stat.qwest.net/flash/current.cfm I don't have an account to see Qwest's ATM status page.

2 1

RE: Faster 'Net growth rate raises fears about routers
by Roeland Meyer 05 Apr '01

05 Apr '01

> From: Sean Donelan [mailto:sean@donelan.com] > Sent: Tuesday, April 03, 2001 5:26 AM > > On Mon, 02 April 2001, Roeland Meyer wrote: > > From the article: > > <quote> > > "Half of the companies that are multihomed should have > gotten better service > > from their providers," says Patrik Faltstrom, a Cisco > engineer and co-chair > > of the IETF's Applications Area. "ISPs haven't done a good > enough job > > explaining to their customers that they don't need to multihome." > > </quote> > > Why would a rational customer pay for a second connection > (usually more than doubling their cost) if a single connection was > satisfactory? Although providers try to vertically integrate > their operations, time and time again, > vertical integration tends to increase the risk for the customer. Because, if the ISP goes out of business the lead-times for getting a new connection exceed 40-days. This was my point on invoking the DSLnetworks/Covad spectre (it's the one I know the most details of). Also reference FlashPoint, NorthPoint, et al. We are now suffering over 100,000 businesses being disconnected from the Intenet because their upstream went tits-up. So, don't tell me it doesn't happen. This is the original reason for having multi-homing with two different upstreams. In most cases, the connection itself was rock-solid up until the time the lights went out. In many cases, there was no warning. In ALL cases, the outage exceeded 4-weeks unless the site was multi-homed somehow. My latest quote for a new DSL connection is 42 days. T1s are 6-8 weeks. Bigger pipes are 2-6 months for the fastest response, many cases are longer. Back to business failures, the DSL world is looking rocky at best. Rythms isn't looking very solid and neither is Covad. Some of the NorthPoint customers are not going with either, electing to go with the local RBOC instead (despite the hassles and long lead-times). What many businesses really need is true multi-homing. > Mid-level providers serve an important function in the > Internet hierarchy. > Multi-homing works well with mid-level providers aggregating > local routes, > and managing redudancy between long-haul providers. If you > don't use a > mid-level provider, to achieve the same reliability you end > up needing to > be your own mid-level provider. Exactly. You also save some monthly recurring cost. > Why can't a large provider operate their network as a set of mid-level > networks, and connections to multiple long-haul networks. They could. But, they dont.

7 7

Re: AS Leakage
by troy＠clownhammer.org 05 Apr '01

05 Apr '01

doubtful jon... they are just practicing that crazy voodoo called "SmartRouting"(tm). doubtful that they wont have other paths to the Mighty UU. Wow, lookey here! They have multiple paths! 4:59pm troy@unagi ~ > host -l customer.alter.net | grep exodus exodus-tco2-gw.customer.alter.net has address 157.130.15.186 peoexodus.customer.alter.net has address 157.130.147.206 exodus-gw.customer.alter.net has address 137.39.133.162 exodus-gw.customer.alter.net has address 146.188.37.78 exodus-gw.customer.alter.net has address 146.188.38.46 exodus-gw.customer.alter.net has address 146.188.38.50 exodus-gw.customer.alter.net has address 146.188.38.54 exodus-lax4-gw.customer.alter.net has address 157.130.12.190 exodus-dca8-oc12-gw.customer.alter.net has address 157.130.60.90 exodus-ewr1-gw.customer.alter.net has address 157.130.12.226 exodus-sea1-oc3.customer.alter.net has address 157.130.12.186 exodus-chi6-gw.customer.alter.net has address 157.130.12.194 exodus-bos1-oc3.customer.alter.net has address 157.130.15.202 exodus-tco1-gw.customer.alter.net has address 157.130.15.182 exodus-tco1-gw.customer.alter.net has address 157.130.12.182 exodus-sjc2-gw.customer.alter.net has address 157.130.12.218 exodus-sjc2-gw.customer.alter.net has address 157.130.12.202 exodus-sjc2-gw.customer.alter.net has address 157.130.12.206 exodus-sjc2-gw.customer.alter.net has address 157.130.12.210 exodus-sjc2-gw.customer.alter.net has address 157.130.12.214 >That's strange. Your 701 connection goes down and you've filtered any >backup route you had into them. Good thinking. > >--- >John Fraizer >EnterZone, Inc > ================================================ thus speaks the Maven of Madness, the Hierophant of Hate, the Minister of Malice, the Rabbi of Rage, the Apothecary of Angst, the Oracle of Outrage, the Vizier of Vituperation, the Archon of Anger... you get the drift. http://www.clownhammer.org

8 13

Re: Faster 'Net growth rate raises fears about routers
by Sean Donelan 05 Apr '01

05 Apr '01

On Tue, 03 April 2001, RJ Atkinson wrote: > At 18:35 03/04/01, Jesper Skriver wrote: > > >Don't know how the world looks like in the US, but here a SDH/Sonet > >provider will never guarentee diversity of his/her circuit to that of a different provider, > > Interesting, though they do guarantee that to NATO > circuits. smd, are you able to get diverse local paths over there ? NATO doesn't get guaranteed diverse circuits. However, during an emergency, NATO does get first dibs on any circuit still working. So, if you have diverse circuits, and one goes down, its possible your other circuit will be pre-empted if it is the only way to restore the NATO circuit.

2 1

RE: Faster 'Net growth rate raises fears about routers
by Roeland Meyer 05 Apr '01

05 Apr '01

> > You can think that way all you want to, right up to the > time when your local > > RBOC decides it will no longer serve you. At which time, > good luck shoving > > those bytes down the wire. > > I may be stupid or something, but the above doesn't appear to make the > least bit of sense. Do you have data that the RBOCs and CLECs > are moving > out of the circuit (DS0 on up) delivery business? No, but individual circuits go down all the time. Simply because you have a big name provider, doesn't mean that they will be more reliable. Only the reasons for the outage change. The recommendations for multi-homing remain the same.

3 2

Major Qwest ATM outage.
by Scott Call 05 Apr '01

05 Apr '01

For any of us unlucky enough to have Qwest ATM service, they are having a major outage. Started exactly at midnight eastern, and the MAST person I spoke with made a reference to lucent conference calls, so I'm betting on a bungled software upgrade or something similar. Qwest Master ticket # 219247. -S Scott Call Router Geek, ATGi "Only the Phoenix rises and does not descend And everything changes. And nothing is truly lost." - Neil Gaiman, The Wake

1 0

Re: Faster 'Net growth rate raises fears about routers
by Sean Donelan 05 Apr '01

05 Apr '01

On Wed, 04 April 2001, Stephen Stuart wrote: > smd(a)clock.org (Hi, Sean) said: > > > P.S.: Isn't it cool that none of us is using an address which in any > > way informs a person who didn't already know, who it is that we > > work for? > > The thing I thought on reading that was that it was further evidence > that you (me, everyody) should be conservative in that you send, and > liberal in what you receive. The comment illustrates (to me, I'm not > trying to put words in Sean's mouth here) that, in an industry as > small as ours, you should exercise caution in dealing with people; > someone working at a startup with a DSL connection to the net today > could very well be an influential person at a global provider or > vendor on which you depend tomorrow. The person you flame tonight > might be the person you need to help you solve a problem tomorrow. I flamed the White House in the Washington Post once. You'll notice I don't post from my employer's account anymore :-)

2 1

Is the answer a bigger club? (was Re: Faster 'Net growth rate raises fears about routers)
by Sean Donelan 05 Apr '01

05 Apr '01

On Wed, 04 April 2001, Stephen Griffin wrote: > Again, very little reason to need multiple providers if the provider > is good. If you're concerned about circuit grooming, write it into > your contract with _severe_ penalties for failure to meet the terms > of the contract. Kung Fu flashback sequence: In time, you will understand. Fortunately, lawyers can not change the laws of nature or statistics no matter what they tell you. No matter what you write into your contract, stuff will go wrong. Car insurance means when you crash your car, the insurance company pays for a new car. It does not mean your car will never crash. By lowering the deductible, all you are changing is the premium you pay. Want a 99% SLA, the premium is $X. Want a 99.999% the premium is $X+N. Want 100% the premium is $X+X+N. VISA (the credit card people) uses two independent providers for its international network, and has a department of people verifying the redundancy of their network. The US Government uses two independent providers for its international network, and has a department of people verifying the redudancy of the circuits which "can't go down." Guess what, NORAD once lost all of its triple-redundant, diversely-routed circuits due to a carrier mistake. If you aren't the size of the US Government or VISA (I'm not sure which has more power) I suspect you won't be able to write penalties more severe than either of them imposed on the carriers. And even with those penalties, circuits still fail due to carrier mistakes. When I did the math, and paid money for circuits, I discovered for the premium I paid for a 100% SLA I could often buy two diverse circuits from two different providers with no SLA for less. And to cap it off, the actual reliability of those two circuits was better than the single 100% SLA circuit I would pay extra.

1 0