June 2001 - Test - lists.nanog.org

Re: Hosting Failover Question
by Paul Froutan 06 Jun '01

06 Jun '01

What do you mean by "reasonable"? Do you mean what you can expect right now at a hoster, or what you should aim for when designing a system? At 6/6/01, James DeMong wrote: >I am looking for a Rule of Thumb on failover for hosting, mostly >tcp(webserver), some udp (audio/video streaming) stuff. >My thought is that 5 seconds to failover when a network >element(switch/router/load balancer/FW) fails is reasonable. > >How many seconds is reasonable for fail over in such a situation? > >Thanks in advance. >__ >James DeMong >Network Design Specialist >TELUS Advanced Communications >Phone: (403) 503-3718 >Email: James.DeMong(a)telus.com ---------------------------------------------------------------------- Paul Froutan Email: pfroutan(a)rackspace.com Rackspace Managed Hosting <http://www.rackspace.com>

1 0

RE: Countdown until UUNET policy affects existing peers
by Barrows, Jeff 06 Jun '01

06 Jun '01

> -----Original Message----- > From: Sean Donelan [mailto:sean@donelan.com] > Sent: Tuesday, June 05, 2001 11:12 PM > To: nanog(a)merit.edu > Subject: Countdown until UUNET policy affects existing peers <snip> > Will UUNET actually carry through on its stated intention to > retroactively apply its new peerin criteria to existing networks > next January? i haven't worked there for a while, and i don't know, but i'd be willing to place a bet that you could count the number of networks that qualify for peering with UUNET (AS 701) today on one finger or less. you make the call. - jsb -- Jeff Barrows SVP, Global Networks Aleron Internet http://www.aleron.com +1 703 287 4224 Office

1 0

RE: HTML-format postings
by Dee McKinney 06 Jun '01

06 Jun '01

I have to use Outlook at the office but it's set to plain text. I'd like to see a rule like that :-) -----Original Message----- From: M. David Leonard [mailto:mdl@equinox.shaysnet.com] Sent: Wednesday, June 06, 2001 10:10 AM To: nanog(a)merit.edu Subject: HTML-format postings Am I the only one who objects to HTML-formatted postings on a listserv? I thought there was a rule that postings should be in plain text. Am I mistaken? David Leonard ShaysNet Curmudgeon-at-large

2 1

RE: C&W Perring
by David Sinn 06 Jun '01

06 Jun '01

So which side is doing the pre-pending? Network Next Hop Metric LocPrf Weight Path * 12.10.44.0/23 166.49.26.73 50 100 0 3561 174 174 174 174 4550 11937 i David -----Original Message----- From: Moe Allen [mailto:moe@vidnet.net] Sent: Wednesday, June 06, 2001 10:33 AM To: Nanog Subject: C&W Perring To all: I just got off the phone with Cable & Wireless. A 60 day continuation of peering with PSINET was reached last night. The peering is back up and we can now find PSINET DNS server again. More on this as I am made aware. Respectfully, Morris Allen President VidcomNet, Inc.

3 2

C&W Perring
by Moe Allen 06 Jun '01

06 Jun '01

To all: I just got off the phone with Cable & Wireless. A 60 day continuation of peering with PSINET was reached last night. The peering is back up and we can now find PSINET DNS server again. More on this as I am made aware. Respectfully, Morris Allen President VidcomNet, Inc.

2 1

RE: C&W/PSI restore connectivity (peering?) Re: C&W Peering
by Barrows, Jeff 06 Jun '01

06 Jun '01

> -----Original Message----- > From: Sean Donelan [mailto:sean@donelan.com] > Sent: Tuesday, June 05, 2001 10:14 PM > To: nanog(a)merit.edu > Subject: C&W/PSI restore connectivity (peering?) Re: C&W Peering <snip> > I would be interested in hearing what folks who once worked as the > "peering coordinator" for a large ISP, and later worked as a smaller > ISP trying to negotiate from different sides of the table think. I > believe every major ISP has turned over their peering staff at least > once, and some multiple times. > > Do the views hold constant no matter where they are, or are they > flexible depending whether they are currently working at the big > ISP or at the startup ISP? An attempt to answer your question without inviting or inciting a debate or detailed discussion [mostly since I couldn't/wouldn't participate] in this forum about the specifics of Worldcom's peering policies or practices: Having been the one responsible for peering for all of Worldcom's IP networks, (including UUNET, ANS, Compuserve, Gridnet, and ~15 other large ASes around the globe that they operate) and having been the one who authored 99% of the most-recently released UUNET peering policy, and having left there to build [from scratch] a global IP backbone which will need direct connectivity to UUNET, I can give you my perspective from both sides of the table. My views on peering, and how an entity should position itself do indeed hold constant. Peering decisions are not technical decisions-- they are business decisions that have an economic impact. No matter what the original intentions were, the major IP networks out there today are all doing business to make money by selling connectivity. How they respond on issues such as peering plays a significant part on their bottom line, and they act accordingly in almost every case. Somewhere close to the heart of the problem is that there must be a line that separates the set of networks that you inter- connect with for free, [or for non-traditional customer pricing,] and the set of networks that you do not... and you'll never get a useful enough sized group of people [that is equally representative of the mix of interests out there] to agree on where that line should be drawn. In the meantime, organizations can draw that line wherever they choose. As this is true for UUNET, Sprint, Genuity, and C&W, it is also true for Aleron. We'll make decisions based on what is best for our customers, our network, and for our business. The position that Aleron will take is different than the position that UUNET will take, even though my views hold constant. Personally, I believe that someone will identify and leverage an opportunity based on all of the market pressures surrounding inter-provider peering, and we'll see a whole new mechanism or method for dealing with "non-customer" inter-network IP traffic exchange. Aleron certainly intends to make great progress in this space. Cheers, - jsb -- Jeff Barrows SVP, Global Networks Aleron Internet http://www.aleron.com +1 703 287 4224 Office

1 0

Re: C&W Peering
by Moe Allen 06 Jun '01

06 Jun '01

Sorry for the delay. Directly from C&W. We have a few Commercial Customers that are unable to connect now with their main office (Lan to Wan). Further investigation and opening a trouble ticket with C&W, produced this information. I generated a letter to peering.admin(a)cwusa.com Asking why and how were our customers going to get back to their main office (via back-haul through the Internet). The reply was less that creative. It will follow later today. Respectfully, Morris Allen VidcomNet, Inc. ----- Original Message ----- From: Jonathan P. Aufderheide To: Moe Allen Sent: Monday, June 04, 2001 3:16 PM Subject: Re: C&W Peering Hi Moe, just out of curiosity, where did you come by this information? thanks, Jonathan P. Aufderheide Director, Global Interconnection Engineering Global Crossing 916.786.8765 jaufder(a)gblx.net ----- Original Message ----- From: Moe Allen To: Nanog Sent: Monday, June 04, 2001 12:41 PM Subject: C&W Peering Affective this morning Cable & Wireless started de-peering with PSINET. More when I receive a reply from: peering.admin(a)cwusa.com Morris Allen VidcomNet, Inc.

1 0

Re: And then there were two
by brandon＠rd.bbc.co.uk 06 Jun '01

06 Jun '01

> Why can't you have more than two 'equals'? Seen Reservoir Dogs? I think it demonstrates the principle regards brandon

1 0

Hosting Failover Question
by James DeMong 06 Jun '01

06 Jun '01

I am looking for a Rule of Thumb on failover for hosting, mostly tcp(webserver), some udp (audio/video streaming) stuff. My thought is that 5 seconds to failover when a network element(switch/router/load balancer/FW) fails is reasonable. How many seconds is reasonable for fail over in such a situation? Thanks in advance. __ James DeMong Network Design Specialist TELUS Advanced Communications Phone: (403) 503-3718 Email: James.DeMong(a)telus.com

1 0

Re: Internet Traffic Discovery?
by Walter Prue 06 Jun '01

06 Jun '01

>Anybody else remember a recent thread arguing that QoS should be kept at >the edge for IP? Well Peter Lothburg (please excuse the spelling) at the last NANOG demonstrated this fact rather well when he showed the effects of congestion on latency for various high speed links. I think he started with T3 then higher and higher up to OC-192. The graphs showed latency low until the links were at progressively higher percentages of 100% full. For links at OC-12 you were looking at links that were well above 90% full before any detectable increase in latency occured. That made the point pretty clear to me. Walt

1 0