On Sun, 3 Jan 1999, Randy Bush wrote:

there has been an interesting discussion on sean's idig list regarding criteria to warrant formal recording. it is an interesting issue. we think we know what we mean by a major outage until we try to formally

This probably doesn't help much in nailing it down.... An outage is major if the general public and the press would perceive it as an Internet outage rather than a single provider outage. A multi-provider outage counts. So does a single provider outage that affects a large enough geographic area that the public would perceive it as multiple outages, i.e. the Internet is down in Boston and New York and all over the place. Even a single provider outage at a backbone that causes outages at several ISP's in a single city would be an "Internet" outage, especially if the backbone were to be one of the dialup modem suppliers to nationals like Earthlink, AOL, et al. I guess the point I am trying to make is to start from public perceptions since we are talking about mainly the reporting of an analysis for publicly known outages. If a SONET ring fails over then there is no publicly known outage. But when it starts to affect multiple points then the public takes note and it is a significant outage that we need to learn from. Could be multiple geographical points, multiple circuits, or multiple providers. -- Michael Dillon - E-mail: michael@memra.com Check the website for my Internet World articles - http://www.memra.com