@Jab said: @VirMach copy-paste from LET, I am to stupid to understand it, but maybe important
@jtk said: I saw reports of intermittent reachability. I suspect, a problem now is that QuadraNet is announcing a /20 in this prefix, but there are only ROAs with Alibaba's origin ASes published in the RPKI for the covered address space, which means many nets will be invalidating and rejecting the QuadraNet announcement. Let's see how long it takes people in the LES thread to figure this out. :-)
That was me. In a nutshell, I suspect some of those who were reporting partial connectivity (e.g., through a VPN, but not from their home cable provider), were able to temporarily reach their systems, because their VPN provider accepted a /20 route from QuadraNet, while their cable provider did not. The reason for some providers not accepting the route was due to a missing, corresponding ROA (route object authorization) published into the RPKI.
In a nutshell, the RPKI is to routing what DNSSEC is for DNS. A ROA is an attestation that a particular origin AS is authorized to announce a particular prefix. If a BGP announcement is made with an origin AS not in a corresponding ROA, many providers will drop the route, and this is known as enforcing ROV (route origin validation). There are now associated ROAs to authorize QuadraNet, among others, to originate prefixes in the /17. This was not the case yesterday.
From what I can tell, Virmach has nothing to do with this process and it is probably all more-or-less automated by Alibaba. My guess is that was probably the last thing Alibaba had to update to restore full connectivity.
@shallow said: Inorite? I'm getting tired of my friends seeing my vps' ip address in our private irc room. I have silly vanity domains that need to be put to use.
That's where IPv6 tunnel come handy :P
Oh, and don't dare to join my channel just to say "Ciao a tutti" (in colours!) and then immediately leave.
Oddly enough after reading most of the thread I had to go over and check them out. Tried to buy a hosting plan, well because reasons, and could not create an account or even buy the vps.
Should I actually contact support or should I just wait a couple months? This server is in Los Angeles; I've got one in Seattle that's still 'on hold' for IPv6, so not sure it's worth bothering support if they're not deploying IPv6 in LA yet.
@ehab said: @VirMach should we open tickets if the vps is missing IP's ? this was from the recent change.
We're working on a solution, we've already flagged down how many IPs each plan should have and will use that to auto clean them up. But you can also use the missing IP button if it's there.
Should I actually contact support or should I just wait a couple months? This server is in Los Angeles; I've got one in Seattle that's still 'on hold' for IPv6, so not sure it's worth bothering support if they're not deploying IPv6 in LA yet.
If it fails and you contact support make sure you make it more about the function not working rather than mentioning you want IPv6 or it will probably be closed/told to use button.
That way we can see what's going on with that so you can make the request.
But at this point the request button isn't really how people think it is, we're basically using it as a voting system to know which ones we should do first in bulk node by node.
(edit) Actually don't even mention the button not working, you should definitely make a ticket for the virtual server missing error as that will cause other problems too.
Comments
Finally, all VPS are back!
Have the honor of being the crybaby who pays $20 for a 128MB VPS at VirMach in 2023.
AMS
LAX
SEA
above are up
SJO seems down
I bench YABS 24/7/365 unless it's a leap year.
That was me. In a nutshell, I suspect some of those who were reporting partial connectivity (e.g., through a VPN, but not from their home cable provider), were able to temporarily reach their systems, because their VPN provider accepted a /20 route from QuadraNet, while their cable provider did not. The reason for some providers not accepting the route was due to a missing, corresponding ROA (route object authorization) published into the RPKI.
In a nutshell, the RPKI is to routing what DNSSEC is for DNS. A ROA is an attestation that a particular origin AS is authorized to announce a particular prefix. If a BGP announcement is made with an origin AS not in a corresponding ROA, many providers will drop the route, and this is known as enforcing ROV (route origin validation). There are now associated ROAs to authorize QuadraNet, among others, to originate prefixes in the /17. This was not the case yesterday.
From what I can tell, Virmach has nothing to do with this process and it is probably all more-or-less automated by Alibaba. My guess is that was probably the last thing Alibaba had to update to restore full connectivity.
Dataplane.org's current server hosting provider list
Patiently waiting to be able to set rDNS for new NL.
It wisnae me! A big boy done it and ran away.
NVMe2G for life! until death (the end is nigh)
GLWRDNS
Haven't bought a single service in VirMach Great Ryzen 2022 - 2023 Flash Sale.
https://lowendspirit.com/uploads/editor/gi/ippw0lcmqowk.png
It's gonna happen... this time.
Inorite? I'm getting tired of my friends seeing my vps' ip address in our private irc room. I have silly vanity domains that need to be put to use.
That's where IPv6 tunnel come handy :P
Oh, and don't dare to join my channel just to say "Ciao a tutti" (in colours!) and then immediately leave.
Weekly check whether Virmach ran away.
No yet, see yor chaps next week
Ontario Dildo Inspector
This is one of the reasons why I check if Virmach has gone out of business once a day.
Hey @VirMach had a decent amount of sleep?
Haven't bought a single service in VirMach Great Ryzen 2022 - 2023 Flash Sale.
https://lowendspirit.com/uploads/editor/gi/ippw0lcmqowk.png
Weed, beers and hookers!
Ontario Dildo Inspector
Laxa032 is timeout right now
I have 2 or 3 vps with extra IP's per vps and after the change have only main IP! did any get the extra IP's visible in the the cp?
it's still time out.
Operation Timed Out After 90001 Milliseconds With 0 Bytes Received
Can't connect via ssh either.
But it shows green on this page: https://statusbot.virm.ac/en/
No one on the same node?
Oddly enough after reading most of the thread I had to go over and check them out. Tried to buy a hosting plan, well because reasons, and could not create an account or even buy the vps.
Should I actually contact support or should I just wait a couple months? This server is in Los Angeles; I've got one in Seattle that's still 'on hold' for IPv6, so not sure it's worth bothering support if they're not deploying IPv6 in LA yet.
TYOC031 is down.
Yeah, just woke up.
rDNS/PTR for Frankfurt, Amsterdam, and Tokyo should be done soon (available on SolusVM.)
These got cleaned up. Not down, just overloading. Ran the newest patch for them yesterday, hopefully will prevent future overloading.
Have you updated the IP Geo info?
https://microlxc.net/
Thanks for quoting my post, I love random notifications.
Haven't bought a single service in VirMach Great Ryzen 2022 - 2023 Flash Sale.
https://lowendspirit.com/uploads/editor/gi/ippw0lcmqowk.png
that's quite a long sleep
I bench YABS 24/7/365 unless it's a leap year.
@VirMach should we open tickets if the vps is missing IP's ? this was from the recent change.
Yes.
https://microlxc.net/
Haven't bought a single service in VirMach Great Ryzen 2022 - 2023 Flash Sale.
https://lowendspirit.com/uploads/editor/gi/ippw0lcmqowk.png
It's been put in, now it's up to them (main few companies) to process it or however long it'll take.
We're working on a solution, we've already flagged down how many IPs each plan should have and will use that to auto clean them up. But you can also use the missing IP button if it's there.
If it fails and you contact support make sure you make it more about the function not working rather than mentioning you want IPv6 or it will probably be closed/told to use button.
That way we can see what's going on with that so you can make the request.
But at this point the request button isn't really how people think it is, we're basically using it as a voting system to know which ones we should do first in bulk node by node.
(edit) Actually don't even mention the button not working, you should definitely make a ticket for the virtual server missing error as that will cause other problems too.