CubeCoders
Failed to login to remote instance null: Two-Factor Login Required
Bug Report
System Information
-
Operating System (Including distribution name and version number) PRETTY_NAME="Ubuntu 22.04.2 LTS" NAME="Ubuntu" VERSION_ID="22.04" VERSION="22.04.2 LTS (Jammy Jellyfish)"
-
AMP version and build date (Always use the version number, 'latest' is not valid!) AMP Release "Decadeus" v2.4.4, built 17/05/2023 17:54
-
Which AMP release stream you're using (Mainline, Nightly or FastTrack) Mainline
I confirm:
- [x] that I have searched for an existing bug report for this issue.
- [x] that I am using the latest available version of AMP.
- [x] that my operating system is up-to-date.
Symptoms
-
What are you trying to do? Navigate through the AMP web interface at my regular speed.
-
What are you expecting to happen? I should be able to navigate through AMP without receiving error messages regardless of the speed at which I click.
-
What is actually happening? If I click through the AMP nav menu too fast I receive this error popup:
Reproduction
Using my exact version of Ubuntu Studio linux, you log into AMP game server and immediately goto an instance. With the new changes, it will load very fast, but once you get into its management screen... if you click on any of the other nav menu items without waiting 10-20 seconds, you will be kicked back to the main ADS01 panel and will receive the 2fa popup error dialogue that states:
Failed to login to remote instance. null Two-Factor Login Required.
After that point, you will have to wait a short time before you may attempt to re-enter the management interface of any of the instances.
##Error logs
this is a section of my ADS log that details the events during the error: [14:06:10] [Logger:tbi Activity] : Managed remote instance UnoriginalIdeaRust-Monthlyvanillanobpwipe01 at http://127.0.0.1:8081/ [14:06:10] [Logger:tbi Activity] : Authentication token for tbi requested by ManageInstance on behalf of tbi [14:07:54] [Logger:tbi Activity] : Managed remote instance UnoriginalIdeaRust-Monthlyvanillanobpwipe01 at http://127.0.0.1:8081/ [14:07:54] [Logger:tbi Activity] : Authentication token for tbi requested by ManageInstance on behalf of tbi [14:09:17] [Logger:tbi Activity] : Managed remote instance UnoriginalIdeaRust-Monthlyvanillanobpwipe01 at http://127.0.0.1:8081/ [14:09:17] [Logger:tbi Activity] : Authentication token for tbi requested by ManageInstance on behalf of tbi [14:09:57] [Logger:tbi Activity] : Managed remote instance UnoriginalIdeaRust-Monthlyvanillanobpwipe01 at http://127.0.0.1:8081/ [14:09:57] [Logger:tbi Activity] : Authentication token for tbi requested by ManageInstance on behalf of tbi [14:10:48] [Logger:tbi Activity] : Managed remote instance UnoriginalIdeaRust-Monthlyvanillanobpwipe01 at http://127.0.0.1:8081/ [14:10:48] [Logger:tbi Activity] : Authentication token for tbi requested by ManageInstance on behalf of tbi [14:24:41] [Logger:tbi Activity] : Managed remote instance UnoriginalIdeaRust-Monthlyvanillanobpwipe01 at http://127.0.0.1:8081/ [14:24:41] [Logger:tbi Activity] : Authentication token for tbi requested by ManageInstance on behalf of tbi
And here is a section of the instances log I was trying to manage... from the same timeframe when the error occurred: [14:24:42] [Logger Activity] : Authentication attempt for user tbi from 127.0.0.1 [14:24:42] [Core Activity] : Authentication success [14:24:44] [Logger:Anonymous Activity] : Authentication attempt for user tbi from 127.0.0.1 [14:24:44] [Logger:Anonymous Activity] : Authentication failure for user tbi from 127.0.0.1 - TwoFactorChallenge
This is still happening on my linux system, after the most recent update, it's actually even worse. Now I cannot even use the "manage" option, I have to manage all my instances in a new tab.
Yup... I'm actually on v2.4.5.4 and still have the issue.
But here is what I figured out... The issue only happens on Chrome. If I fire my panel up in Firefox, the issue goes away instantly.
The latest update to chrome 115.0.5790.171 (Official Build) (64-bit), and the latest update to AMP v2.4.6 (Yeah, I waited a week to update...), it seems to have solved the issue.
Yeah, I was using amp all day at work with it the issue, got home and updated, and now my dev workstation at home works right again.
I will leave this open for now, and monitor things for the next week. If all is good, I will close this.
I have discovered this has something to do with the new Google Password Manager and the "Use Windows Hello when filling passwords" setting.
If turned on, you are no longer troubled by the 2fa popup... the only issue is that you have to put your windows login pin every time you want to fill a password somewhere.
If turned off, you are back to the issue of getting the 2fa popup...
Another update, Previously I had these issues on my Linux chrome installation too.
But upon patching another issue I had with kded5 and docker, it seems to have fixed Chrome and its interactions with AMP as well.
The jist of it is that on my production server, I run a lot of docker applications. What used to happen when I would stop a docker instance was I would get a keyring popup from kded5 on the desktop that would want me to enter my password after the instance stops.
That's all fine and dandy... But why does it want my password when I'm doing unrelated docker things in the terminal. Clearly, that was a bug...
I ended up having to go into "System Settings > KDE Wallet", and unticking "Enable the KDE wallet subsystem", effectively disabling the KDE wallet globally.
So after doing that my actual chrome google password manager stopped syncing with my google account, which is fine, but when I actually log into AMP now... Voila... no more annoying 2fa-popup error when trying to manage instances.
Regarding my Windows development workstation and its Chrome installation, for the time being, I have it set up to prompt for my windows hello pin every time I go to fill in a password. Kind of super annoying to deal with, but w/e.
What's interesting though is that my Windows 10 workstation at work does not actually have the Chrome option to "Use Windows Hello when filling passwords" That workstation does not have a windows hello pin setup, only a password, and the issues still persists on it when I goto manage my amp instances while at work.