Re: Correcting national address databases?

[Nimrod Levy]

Also, address validation in web forms is often "stupid".
Imagine a system for a service that disallows PO boxes.
Now imagine the address you're trying to input is on "Post Office Rd"
NOW imagine trying to explain that to support.
Their solution was to submit a paper form.
My solution was to input "PostOffice Rd" which amazingly worked.

Nimrod


On Thu, May 30, 2024 at 4:31 PM Sean Donelan  wrote:

>
> Since I did address database software for public libraries for a couple of
> decades  Addresses are complicated.
>
> In North American (USA & Canada) there are approximately 80,000
> localities, counties, states and federal addressing authorities (mostly
> local building and planning departments). These are the official (legal)
> addressing authorities.  Tax and real estate records usual the legal
> property definition.
>
>
>
> Because its a PITA to deal with 80,000 different authorities...
>
> The USPS (and Canada Post)) maintains a national directory of addresses
> valid for mailing purposes.  Not every address. The addresses they
> deliver based on local jurisdiction address information, but modified for
> postal needs.
>
> The telephone company (formerly Ma Bell, then ILECs, now PSAPs and states)
> maintain the Master Street Address Guide (MSAG) for E911 purposes. Again
> based on local jurisdication address information, but modified for
> telco/cellular/PSAP needs.
>
> The US Census maintains TIGER and Master Address File (MAF) for planning
> enumeration operations every 10 years. Geocodes domiciles (i.e. where
> people live, not work) for census workers, not necessarily postal or
> telephone addressses.
>
> Several commercial (i.e. expensive) databases for various purposes, such
> as driving, mapping, advertising, etc; all use one of the government
> address databases as their base. Enhance the base information with
> satellites, airplane and street view mapping.
>
>
> When a record is wrong, figuring out the source is a bit of an art.
>
> Mail/Shipping/Ecommerce => start with USPS database
>
> Telco/wireless/E911 => start with MSAG database
>
> Politics/Gerrymandering => start with Census database
>
> Tax, real estate, other => start with local building/planning department
>
>
> On Thu, 30 May 2024, Mike Hammett wrote:
> > This is the Internet, after all, so I will be corrected if I'm wrong.
> > 911 is based on MSAG (Master Street Address Guide), not USPS. However,
> many
> > operators are likely using the USPS system to sanitize the inputs.
>

Re: Correcting national address databases?

[Sean Donelan]

Since I did address database software for public libraries for a couple of 
decades  Addresses are complicated.


In North American (USA & Canada) there are approximately 80,000 
localities, counties, states and federal addressing authorities (mostly 
local building and planning departments). These are the official (legal) 
addressing authorities.  Tax and real estate records usual the legal 
property definition.




Because its a PITA to deal with 80,000 different authorities...

The USPS (and Canada Post)) maintains a national directory of addresses 
valid for mailing purposes.  Not every address. The addresses they 
deliver based on local jurisdiction address information, but modified for 
postal needs.


The telephone company (formerly Ma Bell, then ILECs, now PSAPs and states) 
maintain the Master Street Address Guide (MSAG) for E911 purposes. Again 
based on local jurisdication address information, but modified for 
telco/cellular/PSAP needs.


The US Census maintains TIGER and Master Address File (MAF) for planning 
enumeration operations every 10 years. Geocodes domiciles (i.e. where 
people live, not work) for census workers, not necessarily postal or 
telephone addressses.


Several commercial (i.e. expensive) databases for various purposes, such 
as driving, mapping, advertising, etc; all use one of the government 
address databases as their base. Enhance the base information with 
satellites, airplane and street view mapping.



When a record is wrong, figuring out the source is a bit of an art.

Mail/Shipping/Ecommerce => start with USPS database

Telco/wireless/E911 => start with MSAG database

Politics/Gerrymandering => start with Census database

Tax, real estate, other => start with local building/planning department


On Thu, 30 May 2024, Mike Hammett wrote:

This is the Internet, after all, so I will be corrected if I'm wrong.
911 is based on MSAG (Master Street Address Guide), not USPS. However, many
operators are likely using the USPS system to sanitize the inputs.

Re: Correcting national address databases?

[Andy Brezinsky]

The new standard under NG9-1-1 is GIS with a standard data model. All of 
that is documented by NENA: 
https://cdn.ymaws.com/www.nena.org/resource/resmgr/standards/nena-sta-006.2a_ng9-1-1_gis_.pdf. 
Emergency Call Routing Function (ECRF) uses this data to locate the 
correct PSAP to route a call to.



Some states have their own additions to the NENA standards that aim to 
add consistency to the fields themselves (eg. Post Directional address).



There also exists a standard 
(https://cdn.ymaws.com/www.nena.org/resource/resmgr/Standards/NENA_71-501_GIS_MSAG_ALI_05-.pdf) 
for converting from GIS to MSAG for hand off to non-NG9-1-1 
applications.  It specifies the conversion from PIDF-LO into a MSAG 
record (and the opposite way if needed).




On 5/30/24 14:07, Mike Hammett wrote:

This is the Internet, after all, so I will be corrected if I'm wrong.

911 is based on MSAG (Master Street Address Guide), not USPS. However, 
many operators are likely using the USPS system to sanitize the inputs.




-
Mike Hammett
Intelligent Computing Solutions
http://www.ics-il.com

Midwest-IX
http://www.midwest-ix.com


*From: *"Mike Lewinski via NANOG" 
*To: *nanog@nanog.org
*Sent: *Thursday, May 30, 2024 1:58:20 PM
*Subject: *RE: Correcting national address databases?

On May 30, 2024, at 10:12 AM, Christopher Paul via NANOG 
 wrote:

>
> I propose that there be a national LDAP service, with OUs for each 
zipcode
> (ou=20500,dc=us,dc=gov). A household could register at USPS.gov and 
then be given
> write access to a household OU ("ou=1600 Pennsylvania Ave 
NW,ou=20500,dc=us,dc=gov").
> The household OU could then create inetOrgPersons under that, each 
of which would have

> self-write access.

Your schema is probably good for 99% of the population. I do wonder 
though if USPS is the right / sole agency to maintain. Having 911 
dependent on an incomplete database seems unwise. Or is it ALI? Not 
sure if it was Verizon's front end or back end that was the real 
problem there.


The first time I encountered the problems of living in a place with no 
postal delivery I had a related challenge which was to obtain a new 
driver's license (along with updating vehicle registration and voter 
registration). New Mexico requires two proofs of current residential 
address which for good reasons cannot be a PO Box. The house I was 
renting was fairly new and I don't think USPS knew it even existed. 
There were no road signs or house numbers posted. The first time that 
I visited it the landlord rode along and gave me turn-by-turn 
directions. There was an address on the lease I signed, but I had no 
way to verify it corresponded to the property I visited. In fact later 
I learned that the lease was copied from a template and the address 
had not been updated (when even property owner gets it wrong, what 
hope does a bureaucrat have?)


It took multiple trips to the MVD to obtain a license, being turned 
away several times for insufficient paperwork. I had no utility bills 
for an off-grid home with no postal delivery. In the end they accepted 
a copy of the lease (which I had to photoshop to show the correct 
address) and a statement from the bank. But wait... where did the bank 
get my address? I gave it to them verbally and they accepted it as 
fact. Some time after getting my ID I found a document issued by the 
county that assigned a street address to the house for emergency/law 
enforcement purposes. To my knowledge that is the one and only 
official documentation of the address.


It was around this time (2012) I first became aware of an impending 
REAL ID requirement that the state was rushing to meet. The paradox of 
having to manipulate the system to prove my actual residence to obtain 
a more secure & state-mandated ID card was not lost on me. I never did 
try to update 911 location when I lived there.


This situation of USPS vs 911 vs DMV vs. bank vs. insurer vs. county 
assessor/elections vs. ? reminds me a little of "Gay marriage: the 
database engineering perspective" ( 
https://web.archive.org/web/20170118114056/https://qntm.org/gay ) and 
if I were tasked with creating a grand unified address database all 
those entities could use I'd be studying it and probably also Wes 
Kussmal's "The Sex Life of Tables: What happens when databases about 
you MATE?"


Can I have two entries for two residences? How/who decides which is 
"primary" for income tax purposes then? Can I have zero entries for 
being unhoused but have a cell phone and potentially need 911 
services? If I'm paranoid can I opt-out of that for mental health 
reasons? Can I delete an entry my parents added after I'm disowned, 
preferably without setting up a forwarding entry?


I guess the current state of REAL ID should quash any hopes I have for 
resolving even the relatively simpler problem of 911 USPS location 
dependency.

Re: Correcting national address databases?

[John Levine]

According to Mike Lewinski via NANOG :
>On May 30, 2024, at 10:12 AM, Christopher Paul via NANOG  
>wrote:
>>
>> I propose that there be a national LDAP service, with OUs for each zipcode 
>> (ou=20500,dc=us,dc=gov). A household could register at USPS.gov and then be 
>> given 
>> write access to a household OU ("ou=1600 Pennsylvania Ave 
>> NW,ou=20500,dc=us,dc=gov"). 
>> The household OU could then create inetOrgPersons under that, each of which 
>> would have 
>> self-write access.
>
>Your schema is probably good for 99% of the population. I do wonder though if 
>USPS is the right / sole agency to maintain. Having 911
>dependent on an incomplete database seems unwise. Or is it ALI? Not sure if it 
>was Verizon's front end or back end that was the real
>problem there.

It's hard to imagine that anyone else would do a better job. You're in
an odd and difficult niche which is going to be a niche no matter who
handles it.

When E911 was phased in, some combination of the USPS and local
governments assigned street addresses to everyone who'd prevviously
had an RR or RFD addresss, so "RR#2 Box 27" turned into "473 Pig Burp
Road." I don't see any reason in principle they couldn't add street
addresses for non-deliverable points although there is surely a lot of
bureaucracy to wade through before that could happen.

> The first time I encountered the problems of living in a place with no
> postal delivery I had a related challenge which was to obtain a new
> driver's license (along with updating vehicle registration and voter
> registration). New Mexico requires two proofs of current residential
> address which for good reasons cannot be a PO Box.

You should move to New York. My NY license has always had my PO Box
and no other address.  I do have a street address, and the PO does
deliver there, but it's not on my license.

-- 
Regards,
John Levine, jo...@taugh.com, Primary Perpetrator of "The Internet for Dummies",
Please consider the environment before reading this e-mail. https://jl.ly

Re: Correcting national address databases?

[Mike Hammett]

This is the Internet, after all, so I will be corrected if I'm wrong. 


911 is based on MSAG (Master Street Address Guide), not USPS. However, many 
operators are likely using the USPS system to sanitize the inputs. 




- 
Mike Hammett 
Intelligent Computing Solutions 
http://www.ics-il.com 

Midwest-IX 
http://www.midwest-ix.com 

- Original Message -

From: "Mike Lewinski via NANOG"  
To: nanog@nanog.org 
Sent: Thursday, May 30, 2024 1:58:20 PM 
Subject: RE: Correcting national address databases? 

On May 30, 2024, at 10:12 AM, Christopher Paul via NANOG  
wrote: 
> 
> I propose that there be a national LDAP service, with OUs for each zipcode 
> (ou=20500,dc=us,dc=gov). A household could register at USPS.gov and then be 
> given 
> write access to a household OU ("ou=1600 Pennsylvania Ave 
> NW,ou=20500,dc=us,dc=gov"). 
> The household OU could then create inetOrgPersons under that, each of which 
> would have 
> self-write access. 

Your schema is probably good for 99% of the population. I do wonder though if 
USPS is the right / sole agency to maintain. Having 911 dependent on an 
incomplete database seems unwise. Or is it ALI? Not sure if it was Verizon's 
front end or back end that was the real problem there. 

The first time I encountered the problems of living in a place with no postal 
delivery I had a related challenge which was to obtain a new driver's license 
(along with updating vehicle registration and voter registration). New Mexico 
requires two proofs of current residential address which for good reasons 
cannot be a PO Box. The house I was renting was fairly new and I don't think 
USPS knew it even existed. There were no road signs or house numbers posted. 
The first time that I visited it the landlord rode along and gave me 
turn-by-turn directions. There was an address on the lease I signed, but I had 
no way to verify it corresponded to the property I visited. In fact later I 
learned that the lease was copied from a template and the address had not been 
updated (when even property owner gets it wrong, what hope does a bureaucrat 
have?) 

It took multiple trips to the MVD to obtain a license, being turned away 
several times for insufficient paperwork. I had no utility bills for an 
off-grid home with no postal delivery. In the end they accepted a copy of the 
lease (which I had to photoshop to show the correct address) and a statement 
from the bank. But wait... where did the bank get my address? I gave it to them 
verbally and they accepted it as fact. Some time after getting my ID I found a 
document issued by the county that assigned a street address to the house for 
emergency/law enforcement purposes. To my knowledge that is the one and only 
official documentation of the address. 

It was around this time (2012) I first became aware of an impending REAL ID 
requirement that the state was rushing to meet. The paradox of having to 
manipulate the system to prove my actual residence to obtain a more secure & 
state-mandated ID card was not lost on me. I never did try to update 911 
location when I lived there. 

This situation of USPS vs 911 vs DMV vs. bank vs. insurer vs. county 
assessor/elections vs. ? reminds me a little of "Gay marriage: the database 
engineering perspective" ( 
https://web.archive.org/web/20170118114056/https://qntm.org/gay ) and if I were 
tasked with creating a grand unified address database all those entities could 
use I'd be studying it and probably also Wes Kussmal's "The Sex Life of Tables: 
What happens when databases about you MATE?" 

Can I have two entries for two residences? How/who decides which is "primary" 
for income tax purposes then? Can I have zero entries for being unhoused but 
have a cell phone and potentially need 911 services? If I'm paranoid can I 
opt-out of that for mental health reasons? Can I delete an entry my parents 
added after I'm disowned, preferably without setting up a forwarding entry? 

I guess the current state of REAL ID should quash any hopes I have for 
resolving even the relatively simpler problem of 911 USPS location dependency. 
https://www.theatlantic.com/ideas/archive/2024/05/real-id-deadline-will-never-arrive/678370/
 

RE: Correcting national address databases?

[Mike Lewinski via NANOG]

On May 30, 2024, at 10:12 AM, Christopher Paul via NANOG  
wrote:
>
> I propose that there be a national LDAP service, with OUs for each zipcode 
> (ou=20500,dc=us,dc=gov). A household could register at USPS.gov and then be 
> given 
> write access to a household OU ("ou=1600 Pennsylvania Ave 
> NW,ou=20500,dc=us,dc=gov"). 
> The household OU could then create inetOrgPersons under that, each of which 
> would have 
> self-write access.

Your schema is probably good for 99% of the population. I do wonder though if 
USPS is the right / sole agency to maintain. Having 911 dependent on an 
incomplete database seems unwise. Or is it ALI? Not sure if it was Verizon's 
front end or back end that was the real problem there.

The first time I encountered the problems of living in a place with no postal 
delivery I had a related challenge which was to obtain a new driver's license 
(along with updating vehicle registration and voter registration). New Mexico 
requires two proofs of current residential address which for good reasons 
cannot be a PO Box. The house I was renting was fairly new and I don't think 
USPS knew it even existed. There were no road signs or house numbers posted. 
The first time that I visited it the landlord rode along and gave me 
turn-by-turn directions. There was an address on the lease I signed, but I had 
no way to verify it corresponded to the property I visited. In fact later I 
learned that the lease was copied from a template and the address had not been 
updated (when even property owner gets it wrong, what hope does a bureaucrat 
have?) 

It took multiple trips to the MVD to obtain a license, being turned away 
several times for insufficient paperwork. I had no utility bills for an 
off-grid home with no postal delivery. In the end they accepted a copy of the 
lease (which I had to photoshop to show the correct address) and a statement 
from the bank. But wait... where did the bank get my address? I gave it to them 
verbally and they accepted it as fact. Some time after getting my ID I found a 
document issued by the county that assigned a street address to the house for 
emergency/law enforcement purposes. To my knowledge that is the one and only 
official documentation of the address.

It was around this time (2012) I first became aware of an impending REAL ID 
requirement that the state was rushing to meet. The paradox of having to 
manipulate the system to prove my actual residence to obtain a more secure & 
state-mandated ID card was not lost on me. I never did try to update 911 
location when I lived there.

This situation of USPS vs 911 vs DMV vs. bank vs. insurer vs. county 
assessor/elections vs. ? reminds me a little of "Gay marriage: the database 
engineering perspective" ( 
https://web.archive.org/web/20170118114056/https://qntm.org/gay ) and if I were 
tasked with creating a grand unified address database all those entities could 
use I'd be studying it and probably also Wes Kussmal's "The Sex Life of Tables: 
What happens when databases about you MATE?"

Can I have two entries for two residences? How/who decides which is "primary" 
for income tax purposes then? Can I have zero entries for being unhoused but 
have a cell phone and potentially need 911 services? If I'm paranoid can I 
opt-out of that for mental health reasons? Can I delete an entry my parents 
added after I'm disowned, preferably without setting up a forwarding entry?

I guess the current state of REAL ID should quash any hopes I have for 
resolving even the relatively simpler problem of 911 USPS location dependency.
https://www.theatlantic.com/ideas/archive/2024/05/real-id-deadline-will-never-arrive/678370/

Re: Another outage somewhere in Asia ?

[Grant Taylor via NANOG]

On 5/30/24 6:56 AM, Mehmet wrote:

This is within LA metro


We got bit by this overnight.

Customers transiting our network on the west coast of North America may 
have experienced Latency and Packet Loss. This was due to a fiber cut in 
our vendors network around the Los Angeles area. The vendor has found 
the damaged fiber and have repaired the cables. Cogent NOC is now 
monitoring this event. Customers should have seen their services 
restored at 11:46 AM ET. The master case is HD303782528.


https://ecogent.cogentco.com/network-status




--
Grant. . . .
unix || die

Re: Correcting national address databases?

[Christopher Paul via NANOG]

On 5/30/24 08:53, Mike Lewinski via NANOG wrote:

Another issue is that Amazon (and possibly other online retailers) are charging 
me and my neighbors excess sales tax based on the ZIP code associated with a 
town I do not live in. There's a way to complain and have it reversed for 
every single purchase.

I know this is out of our hands as network operators, but maybe some day one of 
you will be in a position to help.

I propose that there be a national LDAP service, with OUs for each 
zipcode (|ou=20500,dc=us,dc=gov)|. A household could register at 
USPS.gov and then be given write access to a household OU ("ou=1600 
Pennsylvania Ave NW,|ou=20500,dc=us,dc=gov|"). The household OU could 
then create inetOrgPersons under that, each of which would have 
self-write access.


--
Chris Paul | Rex Consulting |https://www.rexconsulting.net

RE: Correcting national address databases?

[Mike Lewinski via NANOG]

That postal database is especially problematic for those who live in rural 
areas with no postal delivery. We need a better database system than the one 
that USPS maintains because it affects a wider range of services.

Two years ago I moved to a house with no postal service, so I  got a PO box in 
town for regular postal mail. I am able to get FedEx, UPS and other non-USPS 
package deliveries to my residence. Google Maps knows right where I live. For 
most purposes, this arrangement is fine. But not all

The first thing I noticed is that Verizon Wireless was unable to update my 911 
location to my physical address because they are pulling from the postal 
database. I tried working through Verizon support a couple different ways but 
no one knew how to fix it. I finally complained to the FCC and after a couple 
months someone at Verizon went in and manually updated my address. I'm sure 
other people are in the same boat. Yeah, hopefully E911 gets a fix in an 
emergency but I'm not counting on it, especially when I'm inside and I see the 
Maps application on my phone putting my location at a nearby cell tower.

The next issue is that it is impossible to apply for credit cards and probably 
other loans online. The banks (more than one I've tried) use the postal 
database and do not recognize my street address and do not accept PO Box input. 

Another issue is that Amazon (and possibly other online retailers) are charging 
me and my neighbors excess sales tax based on the ZIP code associated with a 
town I do not live in. There's a way to complain and have it reversed for 
every single purchase.

I know this is out of our hands as network operators, but maybe some day one of 
you will be in a position to help.



> On May 29, 2024, at 7:14 PM, Aaron C. de Bruyn via NANOG  
> wrote:
> 
> 
> I'm guessing someone in the community has experience dealing with this.
> 
> About 3 years ago my street got typo'd in some sort of national database of 
> addresses.  Two characters were transposed.  i.e. "Mian St" vs "Main St".
> 
> It's causing no end of issues with ordering online, pretty much every shipper 
> has picked up the bad address, and some of the mapping tools too.  Google and 
> OSM appear to be the exceptions.
> 
> Any idea where to go to get this fixed?
> 
> -A

Re: Correcting national address databases?

[Sean Donelan]

Visit your local post office ask for the postmaster.

They can’t fix it but should have a form to correct the database used by almost 
all shippers in the U.S.  unfortunately may take 6 months to propagate changes.

Can also call 1-800-275-8777 but usually local postmaster is helpful.

> On May 29, 2024, at 7:14 PM, Aaron C. de Bruyn via NANOG  
> wrote:
> 
> 
> I'm guessing someone in the community has experience dealing with this.
> 
> About 3 years ago my street got typo'd in some sort of national database of 
> addresses.  Two characters were transposed.  i.e. "Mian St" vs "Main St".
> 
> It's causing no end of issues with ordering online, pretty much every shipper 
> has picked up the bad address, and some of the mapping tools too.  Google and 
> OSM appear to be the exceptions.
> 
> Any idea where to go to get this fixed?
> 
> -A

[NANOG-announce] Get Ready for N91: Q + A with Keynote Kireeti Kompella + MORE!

[Nanog News]

*Get Ready for N91: Q + A with Keynote Kireeti Kompella*
*Juniper Networks' Chief Engineer to Present "Network Digital Twins"*


*"As networks ramp up on automation, this is a logical next step."*
Kompella has been involved off and on with NANOG for nearly two decades. *We
caught up with Kompella to learn more about his upcoming presentation.*

*READ MORE *


*10 More Days Until NANOG 91!*
*Can't Make it to Kansas City? Join us Virtually! *

*In less than two weeks, the NANOG community will gather again!*

Can't make it in-person?

Join us virtually + stream live presentations while engaging in real-time
speaker Q+A plus live chat with other attendees.

*AGENDA + MORE * 


*Take Advantage of NANOG 91 Socials! **Have you Synced the NANOG 91 Socials
to your Calendar?*

Get to know your peers in the industry + foster important relationships in
a relaxed, casual environment.

*GET SOCIAL * 

*Still Time to Register for Hackathon! *
*Hack Your Way into Fun, Friends, + Boost Your Resume! *

*Theme: *Put your sleuthing skills to the test!

*Hybrid Format: *Join us in person in Kansas City or virtually.

*Important Dates:*
Friday, 31 May - Intro/Infrastructure Tutorial/Team Formation
Sunday, 09 June - Competition

*MORE INFO * 
___
NANOG-announce mailing list
NANOG-announce@nanog.org
https://mailman.nanog.org/mailman/listinfo/nanog-announce

Get Ready for N91: Q + A with Keynote Kireeti Kompella + MORE!

[Nanog News]

*Get Ready for N91: Q + A with Keynote Kireeti Kompella*
*Juniper Networks' Chief Engineer to Present "Network Digital Twins"*


*"As networks ramp up on automation, this is a logical next step."*
Kompella has been involved off and on with NANOG for nearly two decades. *We
caught up with Kompella to learn more about his upcoming presentation.*

*READ MORE *


*10 More Days Until NANOG 91!*
*Can't Make it to Kansas City? Join us Virtually! *

*In less than two weeks, the NANOG community will gather again!*

Can't make it in-person?

Join us virtually + stream live presentations while engaging in real-time
speaker Q+A plus live chat with other attendees.

*AGENDA + MORE * 


*Take Advantage of NANOG 91 Socials! **Have you Synced the NANOG 91 Socials
to your Calendar?*

Get to know your peers in the industry + foster important relationships in
a relaxed, casual environment.

*GET SOCIAL * 

*Still Time to Register for Hackathon! *
*Hack Your Way into Fun, Friends, + Boost Your Resume! *

*Theme: *Put your sleuthing skills to the test!

*Hybrid Format: *Join us in person in Kansas City or virtually.

*Important Dates:*
Friday, 31 May - Intro/Infrastructure Tutorial/Team Formation
Sunday, 09 June - Competition

*MORE INFO * 

Re: Another outage somewhere in Asia ?

[Mehmet]

This is within LA metro

Mehmet


On Thu, May 30, 2024 at 07:53 Scott Q.  wrote:

> Oh, ok. That explains why it stops in Lax within Cogent. It appears to be
> a submarine cable cut from their page:
>
> Customers Transiting our network from Europe to our Asian markets may be
> experiencing Latency and Packet Loss. Multiple vessels are currently being
> gathered and/or being dispatched to carry out the repair operations from
> the submarine faults currently being reported. There is NO ETR at this time.
>
>
> On Thursday, 30/05/2024 at 07:27 Mehmet wrote:
>
> I assume this is due to cogent having massive outage in California and
> traffic shifting away from Cogent congesting 1299. It’s my quick
> observation
>
>
> On Thu, May 30, 2024 at 07:24 Scott Q.  wrote:
>
>> Anyone else noticed some major latency / packet loss going to Asia
>> starting roughly 10 hours ago ?
>>
>> For example, tracerouting to 36.68.218.127 provides very mixed results.
>>
>> Traceroute from 192.168.121.126 to 36.68.218.127 (36.68.218.127):
>> 1 192.168.121.1 1.337ms 0.375ms 0.34ms
>> 2 lo0-0.bng4.wh-man.zen.net.uk (51.148.77.131) 5.744ms 4.817ms 4.139ms
>> 3 lag-8.p2.wh-man.zen.net.uk (51.148.73.118) 11.71ms 11.515ms 11.817ms
>> 4 lag-3.p1.thn-lon.zen.net.uk (51.148.73.136) 11.844ms 11.645ms 11.63ms
>> 5 ldn-bb1-link.ip.twelve99.net (62.115.120.74) 12.61ms 12.755ms
>> 12.758ms
>> 6 prs-bb1-link.ip.twelve99.net (62.115.135.25) 19.242ms 18.703ms *
>> 7 mei-b5-link.ip.twelve99.net (62.115.124.55) 28.8ms 28.702ms 28.593ms
>> 8 * prs-bb1-link.ip.twelve99.net (62.115.124.54) 563.155ms *
>> 9 * ldn-bb1-link.ip.twelve99.net (62.115.135.24) 637.818ms *
>> 10 nyk-bb2-link.ip.twelve99.net (62.115.113.20) 364.331ms * *
>> 11 * * *
>> 12 port-b3-link.ip.twelve99.net (62.115.115.25) 1130.258ms 1155.178ms
>> 1149.024ms
>> 13 * tky-b3-link.ip.twelve99.net (62.115.127.11) 1109.197ms *
>> 14 * tky-b2-link.ip.twelve99.net (62.115.126.134) 1498.216ms *
>> 15 * * *
>> 16 * * *
>> 17 * * *
>> 18 180.240.191.166 901.364ms * *
>> 19 * 180.240.204.9 1164.257ms *
>> 20 * * *
>> 21 * * *
>> 22 * * *
>> 23 * * *
>> 24 * * *
>> 255 * * *
>>
>

Re: Another outage somewhere in Asia ?

[Scott Q.]

Oh, ok. That explains why it stops in Lax within Cogent. It
appears to be a submarine cable cut from their page:

Customers Transiting our network from Europe to our Asian markets may
be experiencing Latency and Packet Loss. Multiple vessels are
currently being gathered and/or being dispatched to carry out the
repair operations from the submarine faults currently being reported.
There is NO ETR at this time.

On Thursday, 30/05/2024 at 07:27 Mehmet wrote:




I assume this is due to cogent having massive outage in California and
traffic shifting away from Cogent congesting 1299. It’s my quick
observation 




On Thu, May 30, 2024 at 07:24 Scott Q.  wrote:


 
Anyone else noticed some major latency / packet loss going to Asia
starting roughly 10 hours ago ?

For example, tracerouting to 36.68.218.127 provides very mixed
results.


Traceroute from 192.168.121.126 to 36.68.218.127 (36.68.218.127):
1 192.168.121.1 1.337ms 0.375ms 0.34ms  
2 lo0-0.bng4.wh-man.zen.net.uk [1] (51.148.77.131) 5.744ms 4.817ms
4.139ms  
3 lag-8.p2.wh-man.zen.net.uk [2] (51.148.73.118) 11.71ms 11.515ms
11.817ms  
4 lag-3.p1.thn-lon.zen.net.uk [3] (51.148.73.136) 11.844ms 11.645ms
11.63ms  
5 ldn-bb1-link.ip.twelve99.net [4] (62.115.120.74) 12.61ms 12.755ms
12.758ms  
6 prs-bb1-link.ip.twelve99.net [5] (62.115.135.25) 19.242ms 18.703ms
*  
7 mei-b5-link.ip.twelve99.net [6] (62.115.124.55) 28.8ms 28.702ms
28.593ms  
8 * prs-bb1-link.ip.twelve99.net [5] (62.115.124.54) 563.155ms *  
9 * ldn-bb1-link.ip.twelve99.net [4] (62.115.135.24) 637.818ms *  
10 nyk-bb2-link.ip.twelve99.net [7] (62.115.113.20) 364.331ms * *  
11 * * *  
12 port-b3-link.ip.twelve99.net [8] (62.115.115.25) 1130.258ms
1155.178ms 1149.024ms  
13 * tky-b3-link.ip.twelve99.net [9] (62.115.127.11) 1109.197ms *  
14 * tky-b2-link.ip.twelve99.net [10] (62.115.126.134) 1498.216ms
*  
15 * * *  
16 * * *  
17 * * *  
18 180.240.191.166 901.364ms * *  
19 * 180.240.204.9 1164.257ms *  
20 * * *  
21 * * *  
22 * * *  
23 * * *  
24 * * *  
255 * * *   





Links:
--
[1] http://lo0-0.bng4.wh-man.zen.net.uk
[2] http://lag-8.p2.wh-man.zen.net.uk
[3] http://lag-3.p1.thn-lon.zen.net.uk
[4] http://ldn-bb1-link.ip.twelve99.net
[5] http://prs-bb1-link.ip.twelve99.net
[6] http://mei-b5-link.ip.twelve99.net
[7] http://nyk-bb2-link.ip.twelve99.net
[8] http://port-b3-link.ip.twelve99.net
[9] http://tky-b3-link.ip.twelve99.net
[10] http://tky-b2-link.ip.twelve99.net

Re: Another outage somewhere in Asia ?

[Mehmet]

I assume this is due to cogent having massive outage in California and
traffic shifting away from Cogent congesting 1299. It’s my quick
observation


On Thu, May 30, 2024 at 07:24 Scott Q.  wrote:

> Anyone else noticed some major latency / packet loss going to Asia
> starting roughly 10 hours ago ?
>
> For example, tracerouting to 36.68.218.127 provides very mixed results.
>
> Traceroute from 192.168.121.126 to 36.68.218.127 (36.68.218.127):
> 1 192.168.121.1 1.337ms 0.375ms 0.34ms
> 2 lo0-0.bng4.wh-man.zen.net.uk (51.148.77.131) 5.744ms 4.817ms 4.139ms
> 3 lag-8.p2.wh-man.zen.net.uk (51.148.73.118) 11.71ms 11.515ms 11.817ms
> 4 lag-3.p1.thn-lon.zen.net.uk (51.148.73.136) 11.844ms 11.645ms 11.63ms
> 5 ldn-bb1-link.ip.twelve99.net (62.115.120.74) 12.61ms 12.755ms 12.758ms
> 6 prs-bb1-link.ip.twelve99.net (62.115.135.25) 19.242ms 18.703ms *
> 7 mei-b5-link.ip.twelve99.net (62.115.124.55) 28.8ms 28.702ms 28.593ms
> 8 * prs-bb1-link.ip.twelve99.net (62.115.124.54) 563.155ms *
> 9 * ldn-bb1-link.ip.twelve99.net (62.115.135.24) 637.818ms *
> 10 nyk-bb2-link.ip.twelve99.net (62.115.113.20) 364.331ms * *
> 11 * * *
> 12 port-b3-link.ip.twelve99.net (62.115.115.25) 1130.258ms 1155.178ms
> 1149.024ms
> 13 * tky-b3-link.ip.twelve99.net (62.115.127.11) 1109.197ms *
> 14 * tky-b2-link.ip.twelve99.net (62.115.126.134) 1498.216ms *
> 15 * * *
> 16 * * *
> 17 * * *
> 18 180.240.191.166 901.364ms * *
> 19 * 180.240.204.9 1164.257ms *
> 20 * * *
> 21 * * *
> 22 * * *
> 23 * * *
> 24 * * *
> 255 * * *
>

Another outage somewhere in Asia ?

[Scott Q.]

Anyone else noticed some major latency / packet loss going to
Asia starting roughly 10 hours ago ?

For example, tracerouting to 36.68.218.127 provides very mixed
results.


Traceroute from 192.168.121.126 to 36.68.218.127 (36.68.218.127):
1 192.168.121.1 1.337ms 0.375ms 0.34ms  
2 lo0-0.bng4.wh-man.zen.net.uk (51.148.77.131) 5.744ms 4.817ms
4.139ms  
3 lag-8.p2.wh-man.zen.net.uk (51.148.73.118) 11.71ms 11.515ms
11.817ms  
4 lag-3.p1.thn-lon.zen.net.uk (51.148.73.136) 11.844ms 11.645ms
11.63ms  
5 ldn-bb1-link.ip.twelve99.net (62.115.120.74) 12.61ms 12.755ms
12.758ms  
6 prs-bb1-link.ip.twelve99.net (62.115.135.25) 19.242ms 18.703ms *  
7 mei-b5-link.ip.twelve99.net (62.115.124.55) 28.8ms 28.702ms
28.593ms  
8 * prs-bb1-link.ip.twelve99.net (62.115.124.54) 563.155ms *  
9 * ldn-bb1-link.ip.twelve99.net (62.115.135.24) 637.818ms *  
10 nyk-bb2-link.ip.twelve99.net (62.115.113.20) 364.331ms * *  
11 * * *  
12 port-b3-link.ip.twelve99.net (62.115.115.25) 1130.258ms 1155.178ms
1149.024ms  
13 * tky-b3-link.ip.twelve99.net (62.115.127.11) 1109.197ms *  
14 * tky-b2-link.ip.twelve99.net (62.115.126.134) 1498.216ms *  
15 * * *  
16 * * *  
17 * * *  
18 180.240.191.166 901.364ms * *  
19 * 180.240.204.9 1164.257ms *  
20 * * *  
21 * * *  
22 * * *  
23 * * *  
24 * * *  
255 * * *