From george at ceetonetechnology.com Mon Oct 4 16:15:39 2021 From: george at ceetonetechnology.com (George Rosamond) Date: Mon, 4 Oct 2021 16:15:39 -0400 Subject: [talk] FB outage Message-ID: <53100307-c10a-8ee7-3286-2d9604f89061@ceetonetechnology.com> One of the cooler things we can all learn from the Facebook/Instagram/WA outage is some technical innovations most of you probably missed. https://nitter.eu/briankrebs/status/1445077617426718725 Apparently, there are no RFCs for it, but DNS Global Routing Tables is the place to be. Move over blockchain. g From mwlucas at michaelwlucas.com Mon Oct 4 17:17:59 2021 From: mwlucas at michaelwlucas.com (Michael W. Lucas) Date: Mon, 4 Oct 2021 17:17:59 -0400 Subject: [talk] FB outage In-Reply-To: <53100307-c10a-8ee7-3286-2d9604f89061@ceetonetechnology.com> References: <53100307-c10a-8ee7-3286-2d9604f89061@ceetonetechnology.com> Message-ID: That's it. Twitter is the next to go. ==ml On Mon, Oct 04, 2021 at 04:15:39PM -0400, George Rosamond wrote: > One of the cooler things we can all learn from the Facebook/Instagram/WA > outage is some technical innovations most of you probably missed. > > https://nitter.eu/briankrebs/status/1445077617426718725 > > Apparently, there are no RFCs for it, but DNS Global Routing Tables is > the place to be. Move over blockchain. > > g > > _______________________________________________ > talk mailing list > talk at lists.nycbug.org > http://lists.nycbug.org:8080/mailman/listinfo/talk -- Michael W. Lucas https://mwl.io/ author of: Absolute OpenBSD, SSH Mastery, git commit murder, Absolute FreeBSD, Immortal Clay, Prohibition Orcs, etc, etc, etc... ### New books: TLS Mastery, the Networknomicon, $ git sync murder ### From imp at bsdimp.com Mon Oct 4 17:28:13 2021 From: imp at bsdimp.com (Warner Losh) Date: Mon, 4 Oct 2021 15:28:13 -0600 Subject: [talk] FB outage In-Reply-To: References: <53100307-c10a-8ee7-3286-2d9604f89061@ceetonetechnology.com> Message-ID: The rumors from my side. I've not verified them all (some are in the twitter feed). We'll see how reliable this is... But, Facebook does all their own tooling, and it's all down. Even internal chat. So they are talking directly via cell phones (system to lookup cell numbers, also down) The only way that can provision the bgp routes is via the console of the routers at this point... ... but the internal system to get credentials is also down. They also need to get someone in front of the console to bring things back online... Their tools to coordinate and meet to come up with a plan, also down.... Single point of failure in the extreme. On Mon, Oct 4, 2021 at 3:20 PM Michael W. Lucas wrote: > > > > That's it. Twitter is the next to go. > > ==ml > > > On Mon, Oct 04, 2021 at 04:15:39PM -0400, George Rosamond wrote: > > One of the cooler things we can all learn from the Facebook/Instagram/WA > > outage is some technical innovations most of you probably missed. > > > > https://nitter.eu/briankrebs/status/1445077617426718725 > > > > Apparently, there are no RFCs for it, but DNS Global Routing Tables is > > the place to be. Move over blockchain. > > > > g > > > > _______________________________________________ > > talk mailing list > > talk at lists.nycbug.org > > http://lists.nycbug.org:8080/mailman/listinfo/talk > > -- > Michael W. Lucas https://mwl.io/ > author of: Absolute OpenBSD, SSH Mastery, git commit murder, > Absolute FreeBSD, Immortal Clay, Prohibition Orcs, etc, etc, etc... > ### New books: TLS Mastery, the Networknomicon, $ git sync murder ### > > _______________________________________________ > talk mailing list > talk at lists.nycbug.org > http://lists.nycbug.org:8080/mailman/listinfo/talk > -------------- next part -------------- An HTML attachment was scrubbed... URL: From jim at netgate.com Mon Oct 4 17:35:48 2021 From: jim at netgate.com (Jim Thompson) Date: Mon, 4 Oct 2021 15:35:48 -0600 Subject: [talk] FB outage In-Reply-To: References: <53100307-c10a-8ee7-3286-2d9604f89061@ceetonetechnology.com> Message-ID: <29668ECA-820D-4271-819E-47F63B5CE474@netgate.com> Rumor has it that even the card/badge access systems are down. https://twitter.com/sheeraf/status/1445099150316503057 So getting someone in-front of a console is ? challenging. > On Oct 4, 2021, at 3:28 PM, Warner Losh wrote: > > The rumors from my side. I've not verified them all (some are in the twitter feed). We'll see how reliable this is... > > But, Facebook does all their own tooling, and it's all down. > Even internal chat. So they are talking directly via cell phones (system to lookup cell numbers, also down) > The only way that can provision the bgp routes is via the console of the routers at this point... > ... but the internal system to get credentials is also down. > They also need to get someone in front of the console to bring things back online... > Their tools to coordinate and meet to come up with a plan, also down.... > > Single point of failure in the extreme. > > > On Mon, Oct 4, 2021 at 3:20 PM Michael W. Lucas > wrote: > > > > That's it. Twitter is the next to go. > > ==ml > > > On Mon, Oct 04, 2021 at 04:15:39PM -0400, George Rosamond wrote: > > One of the cooler things we can all learn from the Facebook/Instagram/WA > > outage is some technical innovations most of you probably missed. > > > > https://nitter.eu/briankrebs/status/1445077617426718725 > > > > Apparently, there are no RFCs for it, but DNS Global Routing Tables is > > the place to be. Move over blockchain. > > > > g > > > > _______________________________________________ > > talk mailing list > > talk at lists.nycbug.org > > http://lists.nycbug.org:8080/mailman/listinfo/talk > > -- > Michael W. Lucas https://mwl.io/ > author of: Absolute OpenBSD, SSH Mastery, git commit murder, > Absolute FreeBSD, Immortal Clay, Prohibition Orcs, etc, etc, etc... > ### New books: TLS Mastery, the Networknomicon, $ git sync murder ### > > _______________________________________________ > talk mailing list > talk at lists.nycbug.org > http://lists.nycbug.org:8080/mailman/listinfo/talk > _______________________________________________ > talk mailing list > talk at lists.nycbug.org > https://www.google.com/url?q=http://lists.nycbug.org:8080/mailman/listinfo/talk&source=gmail-imap&ust=1633987767000000&usg=AOvVaw0m3zHyt7BS3Bgj3MP_B1en -------------- next part -------------- An HTML attachment was scrubbed... URL: From pete at nomadlogic.org Mon Oct 4 17:38:35 2021 From: pete at nomadlogic.org (Pete Wright) Date: Mon, 4 Oct 2021 14:38:35 -0700 Subject: [talk] FB outage In-Reply-To: <29668ECA-820D-4271-819E-47F63B5CE474@netgate.com> References: <53100307-c10a-8ee7-3286-2d9604f89061@ceetonetechnology.com> <29668ECA-820D-4271-819E-47F63B5CE474@netgate.com> Message-ID: <87f8c354-2671-a12f-d0b8-0675e4b5c018@nomadlogic.org> On 10/4/21 2:35 PM, Jim Thompson wrote: > Rumor has it that even the card/badge access systems are down. > https://twitter.com/sheeraf/status/1445099150316503057 > > > So getting someone in-front of a console is ? challenging. it'll be interesting to see how many people need to add *this* to their tabletop exercises this year! -p -- Pete Wright pete at nomadlogic.org @nomadlogicLA From jondrews at fastmail.com Mon Oct 4 19:31:20 2021 From: jondrews at fastmail.com (Jonathan Drews) Date: Mon, 04 Oct 2021 17:31:20 -0600 Subject: [talk] FB outage In-Reply-To: <53100307-c10a-8ee7-3286-2d9604f89061@ceetonetechnology.com> References: <53100307-c10a-8ee7-3286-2d9604f89061@ceetonetechnology.com> Message-ID: <3248365f-b394-4c28-86a2-27d6515abcd8@www.fastmail.com> "Maybe Zuck got mad that he kept getting roasted for running a junk company and he rage quit the internet lol" A quote from the link you provided George. On Mon, Oct 4, 2021, at 14:15, George Rosamond wrote: > One of the cooler things we can all learn from the Facebook/Instagram/WA > outage is some technical innovations most of you probably missed. > > https://nitter.eu/briankrebs/status/1445077617426718725 > > Apparently, there are no RFCs for it, but DNS Global Routing Tables is > the place to be. Move over blockchain. > > g > > _______________________________________________ > talk mailing list > talk at lists.nycbug.org > http://lists.nycbug.org:8080/mailman/listinfo/talk > -------------- next part -------------- An HTML attachment was scrubbed... URL: From steve.b at osfda.org Mon Oct 4 19:35:30 2021 From: steve.b at osfda.org (steve.b at osfda.org) Date: Mon, 4 Oct 2021 19:35:30 -0400 Subject: [talk] FB outage In-Reply-To: References: <53100307-c10a-8ee7-3286-2d9604f89061@ceetonetechnology.com> Message-ID: <324664c1-f961-0d46-5538-47e854a4ad8b@osfda.org> CTO recently left. If it's bgp, happens to all of them... On 10/4/21 5:28 PM, Warner Losh wrote: > The rumors from my side. I've not verified them all (some are in the > twitter feed). We'll see how reliable this is... > > But, Facebook does all their own tooling, and it's all down. > Even internal chat. So they are talking directly via cell phones > (system to lookup cell numbers, also down) > The only way that can provision the bgp routes is via the console of > the routers at this point... > ... but the internal system to get credentials?is also down. > They also need to get someone in front of the console to bring things > back online... > Their tools to coordinate and meet to come up with a plan, also down.... > > Single point of failure in the extreme. > > > On Mon, Oct 4, 2021 at 3:20 PM Michael W. Lucas > > wrote: > > > > > That's it. Twitter is the next to go. > > ==ml > > > On Mon, Oct 04, 2021 at 04:15:39PM -0400, George Rosamond wrote: > > One of the cooler things we can all learn from the > Facebook/Instagram/WA > > outage is some technical innovations most of you probably missed. > > > > https://nitter.eu/briankrebs/status/1445077617426718725 > > > > > Apparently, there are no RFCs for it, but DNS Global Routing > Tables is > > the place to be. Move over blockchain. > > > > g > > > > _______________________________________________ > > talk mailing list > > talk at lists.nycbug.org > > http://lists.nycbug.org:8080/mailman/listinfo/talk > > > -- > Michael W. Lucas https://mwl.io/ > author of: Absolute OpenBSD, SSH Mastery, git commit murder, > ?Absolute FreeBSD, Immortal Clay, Prohibition Orcs, etc, etc, etc... > ### New books: TLS Mastery, the Networknomicon, $ git sync murder ### > > _______________________________________________ > talk mailing list > talk at lists.nycbug.org > http://lists.nycbug.org:8080/mailman/listinfo/talk > > > > _______________________________________________ > talk mailing list > talk at lists.nycbug.org > http://lists.nycbug.org:8080/mailman/listinfo/talk -------------- next part -------------- An HTML attachment was scrubbed... URL: From george at ceetonetechnology.com Mon Oct 4 19:44:31 2021 From: george at ceetonetechnology.com (George Rosamond) Date: Mon, 4 Oct 2021 19:44:31 -0400 Subject: [talk] FB outage In-Reply-To: <324664c1-f961-0d46-5538-47e854a4ad8b@osfda.org> References: <53100307-c10a-8ee7-3286-2d9604f89061@ceetonetechnology.com> <324664c1-f961-0d46-5538-47e854a4ad8b@osfda.org> Message-ID: <3876bc93-e0b0-3c7f-b429-a0bc4dc43103@ceetonetechnology.com> On 10/4/21 19:35, steve.b at osfda.org wrote: > CTO recently left. > > If it's bgp, happens to all of them... DNS was the first diagnois, but I tested the .onion site over Tor (ie "dark web") and once it wasn't accessible it became clear it wasn't DNS. I have no idea about the procedures and processes on editing BGP routing tables at a FB or wherever, but it seems a little crazy that you can have a single-point of failure like one wily/confused/hung-over Yankee fan could fat finger with an impact like that. Not necessarily literally, but you get my point... Jan noted in a tweet that over 12000 domains used the Facebook DNS as auth. Am I the only one who was distracted by reading and chatting about it all day? g From pvarga at pvrg.net Mon Oct 4 19:46:37 2021 From: pvarga at pvrg.net (Peter Varga) Date: Mon, 04 Oct 2021 16:46:37 -0700 Subject: [talk] FB outage In-Reply-To: <324664c1-f961-0d46-5538-47e854a4ad8b@osfda.org> References: <53100307-c10a-8ee7-3286-2d9604f89061@ceetonetechnology.com> <324664c1-f961-0d46-5538-47e854a4ad8b@osfda.org> Message-ID: <822c7b00-5d73-4a76-a390-f370f873e5fa@www.fastmail.com> It is bgp also On Mon, Oct 4, 2021, at 16:35, steve.b at osfda.org wrote: > CTO recently left. > > If it's bgp, happens to all of them... > > On 10/4/21 5:28 PM, Warner Losh wrote: >> The rumors from my side. I've not verified them all (some are in the twitter feed). We'll see how reliable this is... >> >> But, Facebook does all their own tooling, and it's all down. >> Even internal chat. So they are talking directly via cell phones (system to lookup cell numbers, also down) >> The only way that can provision the bgp routes is via the console of the routers at this point... >> ... but the internal system to get credentials is also down. >> They also need to get someone in front of the console to bring things back online... >> Their tools to coordinate and meet to come up with a plan, also down.... >> >> Single point of failure in the extreme. >> >> >> On Mon, Oct 4, 2021 at 3:20 PM Michael W. Lucas wrote: >>> >>> >>> >>> That's it. Twitter is the next to go. >>> >>> ==ml >>> >>> >>> On Mon, Oct 04, 2021 at 04:15:39PM -0400, George Rosamond wrote: >>> > One of the cooler things we can all learn from the Facebook/Instagram/WA >>> > outage is some technical innovations most of you probably missed. >>> > >>> > https://nitter.eu/briankrebs/status/1445077617426718725 >>> > >>> > Apparently, there are no RFCs for it, but DNS Global Routing Tables is >>> > the place to be. Move over blockchain. >>> > >>> > g >>> > >>> > _______________________________________________ >>> > talk mailing list >>> > talk at lists.nycbug.org >>> > http://lists.nycbug.org:8080/mailman/listinfo/talk >>> >>> -- >>> Michael W. Lucas https://mwl.io/ >>> author of: Absolute OpenBSD, SSH Mastery, git commit murder, >>> Absolute FreeBSD, Immortal Clay, Prohibition Orcs, etc, etc, etc... >>> ### New books: TLS Mastery, the Networknomicon, $ git sync murder ### >>> >>> _______________________________________________ >>> talk mailing list >>> talk at lists.nycbug.org >>> http://lists.nycbug.org:8080/mailman/listinfo/talk >> >> _______________________________________________ >> talk mailing list >> talk at lists.nycbug.org >> http://lists.nycbug.org:8080/mailman/listinfo/talk >> > _______________________________________________ > talk mailing list > talk at lists.nycbug.org > http://lists.nycbug.org:8080/mailman/listinfo/talk > -------------- next part -------------- An HTML attachment was scrubbed... URL: From steve.b at osfda.org Mon Oct 4 19:52:11 2021 From: steve.b at osfda.org (steve.b at osfda.org) Date: Mon, 4 Oct 2021 19:52:11 -0400 Subject: [talk] FB outage In-Reply-To: <324664c1-f961-0d46-5538-47e854a4ad8b@osfda.org> References: <53100307-c10a-8ee7-3286-2d9604f89061@ceetonetechnology.com> <324664c1-f961-0d46-5538-47e854a4ad8b@osfda.org> Message-ID: <7fe1def8-1c07-2815-2efc-dbae226cb281@osfda.org> Akami, Google, DoD -they all have had serious bgp outages due to config mistakes. IF that's what it was... On 10/4/21 7:35 PM, steve.b at osfda.org wrote: > > CTO recently left. > > If it's bgp, happens to all of them... > > On 10/4/21 5:28 PM, Warner Losh wrote: >> The rumors from my side. I've not verified them all (some are in the >> twitter feed). We'll see how reliable this is... >> >> But, Facebook does all their own tooling, and it's all down. >> Even internal chat. So they are talking directly via cell phones >> (system to lookup cell numbers, also down) >> The only way that can provision the bgp routes is via the console of >> the routers at this point... >> ... but the internal system to get credentials?is also down. >> They also need to get someone in front of the console to bring things >> back online... >> Their tools to coordinate and meet to come up with a plan, also down.... >> >> Single point of failure in the extreme. >> >> >> On Mon, Oct 4, 2021 at 3:20 PM Michael W. Lucas >> > wrote: >> >> >> >> >> That's it. Twitter is the next to go. >> >> ==ml >> >> >> On Mon, Oct 04, 2021 at 04:15:39PM -0400, George Rosamond wrote: >> > One of the cooler things we can all learn from the >> Facebook/Instagram/WA >> > outage is some technical innovations most of you probably missed. >> > >> > https://nitter.eu/briankrebs/status/1445077617426718725 >> >> > >> > Apparently, there are no RFCs for it, but DNS Global Routing >> Tables is >> > the place to be. Move over blockchain. >> > >> > g >> > >> > _______________________________________________ >> > talk mailing list >> > talk at lists.nycbug.org >> > http://lists.nycbug.org:8080/mailman/listinfo/talk >> >> >> -- >> Michael W. Lucas https://mwl.io/ >> author of: Absolute OpenBSD, SSH Mastery, git commit murder, >> ?Absolute FreeBSD, Immortal Clay, Prohibition Orcs, etc, etc, etc... >> ### New books: TLS Mastery, the Networknomicon, $ git sync murder ### >> >> _______________________________________________ >> talk mailing list >> talk at lists.nycbug.org >> http://lists.nycbug.org:8080/mailman/listinfo/talk >> >> >> >> _______________________________________________ >> talk mailing list >> talk at lists.nycbug.org >> http://lists.nycbug.org:8080/mailman/listinfo/talk -------------- next part -------------- An HTML attachment was scrubbed... URL: From george at ceetonetechnology.com Mon Oct 4 19:59:20 2021 From: george at ceetonetechnology.com (George Rosamond) Date: Mon, 4 Oct 2021 19:59:20 -0400 Subject: [talk] FB outage In-Reply-To: <7fe1def8-1c07-2815-2efc-dbae226cb281@osfda.org> References: <53100307-c10a-8ee7-3286-2d9604f89061@ceetonetechnology.com> <324664c1-f961-0d46-5538-47e854a4ad8b@osfda.org> <7fe1def8-1c07-2815-2efc-dbae226cb281@osfda.org> Message-ID: <36dfde8c-4ff1-78df-170c-fb65d2216dde@ceetonetechnology.com> On 10/4/21 19:52, steve.b at osfda.org wrote: > Akami, Google, DoD -they all have had serious bgp outages due to config > mistakes. Yes, we all probably remember Pakistan Telecom... https://www.wired.com/2008/02/pakistans-accid/ Then Alex P and a friend turned it into free tickets to DefCon: https://www.wired.com/2008/08/revealed-the-in/ > > IF that's what it was... Sure, we really have no idea since there's been nothing "official" that is totally trustable and believable from a company that is clearly neither. But the reality is that the BGP issues are verifiable... g From ike at blackskyresearch.net Mon Oct 4 20:54:40 2021 From: ike at blackskyresearch.net (Isaac (.ike) Levy) Date: Mon, 4 Oct 2021 20:54:40 -0400 Subject: [talk] FB outage In-Reply-To: <36dfde8c-4ff1-78df-170c-fb65d2216dde@ceetonetechnology.com> References: <53100307-c10a-8ee7-3286-2d9604f89061@ceetonetechnology.com> <324664c1-f961-0d46-5538-47e854a4ad8b@osfda.org> <7fe1def8-1c07-2815-2efc-dbae226cb281@osfda.org> <36dfde8c-4ff1-78df-170c-fb65d2216dde@ceetonetechnology.com> Message-ID: > On Oct 4, 2021, at 7:59 PM, George Rosamond wrote: > > On 10/4/21 19:52, steve.b at osfda.org wrote: >> Akami, Google, DoD -they all have had serious bgp outages due to config mistakes. > > Yes, we all probably remember Pakistan Telecom... > > https://www.wired.com/2008/02/pakistans-accid/ > > Then Alex P and a friend turned it into free tickets to DefCon: > https://www.wired.com/2008/08/revealed-the-in/ > >> IF that's what it was... > > Sure, we really have no idea since there's been nothing "official" that is totally trustable and believable from a company that is clearly neither. > > But the reality is that the BGP issues are verifiable... > > g But what is FaceBook? Rocket- .ike From raulcuza at gmail.com Mon Oct 4 21:09:06 2021 From: raulcuza at gmail.com (Raul Cuza) Date: Mon, 4 Oct 2021 21:09:06 -0400 Subject: [talk] FB outage In-Reply-To: References: <53100307-c10a-8ee7-3286-2d9604f89061@ceetonetechnology.com> <324664c1-f961-0d46-5538-47e854a4ad8b@osfda.org> <7fe1def8-1c07-2815-2efc-dbae226cb281@osfda.org> <36dfde8c-4ff1-78df-170c-fb65d2216dde@ceetonetechnology.com> Message-ID: On Mon, Oct 4, 2021 at 20:55 Isaac (.ike) Levy wrote: > > > On Oct 4, 2021, at 7:59 PM, George Rosamond < > george at ceetonetechnology.com> wrote: > > > > On 10/4/21 19:52, steve.b at osfda.org wrote: > >> Akami, Google, DoD -they all have had serious bgp outages due to config > mistakes. > > > > Yes, we all probably remember Pakistan Telecom... > > > > https://www.wired.com/2008/02/pakistans-accid/ > > > > Then Alex P and a friend turned it into free tickets to DefCon: > > https://www.wired.com/2008/08/revealed-the-in/ > > > >> IF that's what it was... > > > > Sure, we really have no idea since there's been nothing "official" that > is totally trustable and believable from a company that is clearly neither. > > > > But the reality is that the BGP issues are verifiable... > > > > g > > But what is FaceBook? > > Rocket- > .ike What if nobody noticed? Maybe in BGP v6 we can build in a stochastic feature that turns off the Internet for an hour. If the number of people who die from that feature is low enough, it would make the human part of the system more resilient. - r > -------------- next part -------------- An HTML attachment was scrubbed... URL: From greg at ltcc.com Mon Oct 4 22:46:05 2021 From: greg at ltcc.com (Greg Robinson) Date: Mon, 4 Oct 2021 19:46:05 -0700 Subject: [talk] FB outage In-Reply-To: <53100307-c10a-8ee7-3286-2d9604f89061@ceetonetechnology.com> References: <53100307-c10a-8ee7-3286-2d9604f89061@ceetonetechnology.com> Message-ID: <004501d7b993$25ebcb20$71c36160$@ltcc.com> Frankly, it's a welcome event. Best, Greg Lucrosol 415.922.3675 (PST) ?You get rid of the thought of competition. You are to create, not to compete for what is already created. You do not have to take anything away from any?one? Wallace D. Wattles -----Original Message----- From: talk [mailto:talk-bounces at lists.nycbug.org] On Behalf Of George Rosamond Sent: Monday, October 04, 2021 1:16 PM To: talk Subject: [talk] FB outage One of the cooler things we can all learn from the Facebook/Instagram/WA outage is some technical innovations most of you probably missed. https://nitter.eu/briankrebs/status/1445077617426718725 Apparently, there are no RFCs for it, but DNS Global Routing Tables is the place to be. Move over blockchain. g _______________________________________________ talk mailing list talk at lists.nycbug.org http://lists.nycbug.org:8080/mailman/listinfo/talk From jondrews at fastmail.com Mon Oct 18 20:14:55 2021 From: jondrews at fastmail.com (Jonathan Drews) Date: Mon, 18 Oct 2021 18:14:55 -0600 Subject: [talk] Semi-BUG Presentation on EMACS Message-ID: Hi Folks: SEMI-Bug will host a presentation on EMACS tomorrow, Tuesday, October, 19th, at 1900 Detroit, MI time. The Jitsi URL is: https://meet.jit.si/SEMI-BugEMACSPresentation -- Kind regards, Jonathan