No longer able to log in to pve - no GUI authentication pop up appears

O

oldtimer

New Member
Feb 2, 2024
11
0
1
Really don't know what to say about this. Overnight I have become unable to log into one of the pve servers in my cluster. When I try I just get a blank screen without any login popup.

Both the other macines in the cluster work fine and I have full control over the 3rd machine to which I am unable to log in. No idea where to even start looking for this.

I could really do with some guidance on this. TIA.

probably should add that I can also migrate, start, stop any VM.
 
Last edited:
Have you checked your browser console for errors? Have you disabled any password managers or other extensions that may interfere? Have you previously installed any "enhancements" that affect non-production subscription reminder?

F

[SOLVED] Thread 'Lost acess to Desktop GUIin proxmox'

I'm facing a strange issue with my Proxmox cluster and would appreciate some guidance. After a recent update on two of my nodes, the desktop version of the web GUI is blank, while the mobile version works perfectly.

Background / Recent Changes:

I have a cluster of several Proxmox servers.

Yesterday, I ran apt update followed by apt upgrade on two of the nodes. I was working via SSH and didn't check the GUI at that time.

Today, I can no longer access the desktop Web GUI on those two updated nodes.

Crucially, one node in the cluster that I did not update still has a perfectly working...


Blockbridge : Ultra low latency all-NVME shared storage for Proxmox - https://www.blockbridge.com/proxmox
 
I tried with several different browsers - including in private mode, and the other two PVE managers in the cluster work perfectly and allow me access to the problem one.

password manager switched off, same response

the console is not showing up anything that I can see

I actually have paid subscriptions for the two main PVEs in the cluster because I want to support the project (the third is just to enable a quorum using a low power PC as the second one is a backup mostly turned off). it is the main pve machine that is exhibiting this behavior. I have tried updating and rebooting to no avail.

In any event, I am not getting as far as logging onto the pve manager on that machine. When the sign-in pop up should appear I just have a black windoe (in dark mode).
 
No, but having a quick look at it sounds like I have the same issues. I will work through that. I did do a search, but obviously my search terms didn't catch this thread. Thanks for the link. I'll get round to it later today and report back tomorrow.
 
Back again Onslow. I have worked through everything and gotten the exact same result.

In case it helps, her is the curl response suggested in the other thread

Code: 
~# curl -k https://172.16.1.15:8006
<!DOCTYPE html>
<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
    <meta http-equiv="X-UA-Compatible" content="IE=edge">
    <meta name="viewport" content="width=device-width, initial-scale=1, maximum-scale=1, user-scalable=no">
    <title>vm1 - Proxmox Virtual Environment</title>
    <link rel="icon" sizes="128x128" href="/pve2/images/logo-128.png" />
    <link rel="apple-touch-icon" sizes="128x128" href="/pve2/images/logo-128.png" />
    <link rel="stylesheet" type="text/css" href="/pve2/ext6/theme-crisp/resources/theme-crisp-all.css?ver=7.0.0" />
    <link rel="stylesheet" type="text/css" href="/pve2/ext6/crisp/resources/charts-all.css?ver=7.0.0" />
    <link rel="stylesheet" type="text/css" href="/pve2/fa/css/font-awesome.css" />
    <link rel="stylesheet" type="text/css" href="/pve2/font-logos/css/font-logos.css" />
    <link rel="stylesheet" type="text/css" href="/pve2/css/ext6-pve.css?ver=8.4.14" />
    <link rel="stylesheet" type="text/css" href="/pwt/css/ext6-pmx.css?ver=v4.3.13-t1754315429" />
    <link rel="stylesheet" type="text/css" media="(prefers-color-scheme: dark)" href="/pwt/themes/theme-proxmox-dark.css?ver=v4.3.13-t1754315429" />
    
    <script type='text/javascript'>function gettext(buf) { return buf; }</script>
    
    <script type="text/javascript" src="/pve2/ext6/ext-all.js?ver=7.0.0"></script>
    <script type="text/javascript" src="/pve2/ext6/charts.js?ver=7.0.0"></script>
    
    <script type="text/javascript" src="/pve2/js/u2f-api.js"></script>
    <script type="text/javascript" src="/qrcode.min.js"></script>
    <script type="text/javascript">
    Proxmox = {
    Setup: { auth_cookie_name: 'PVEAuthCookie' },
    defaultLang: 'en',
    NodeName: 'vm1',
    UserName: '',
    CSRFPreventionToken: 'null',
    ConsentText: ''
    };
    </script>
    <script type="text/javascript" src="/proxmoxlib.js?ver=v4.3.13-t1754315429"></script>
    <script type="text/javascript" src="/pve2/js/pvemanagerlib.js?ver=8.4.14"></script>
    <script type="text/javascript" src="/pve2/ext6/locale/locale-en.js?ver=7.0.0"></script>

    <script type="text/javascript">
    if (typeof(PVE) === 'undefined') PVE = {};
    Ext.History.fieldid = 'x-history-field';
    Ext.onReady(function() { Ext.create('PVE.StdWorkspace');});
    </script>

  </head>
  <body>
    <!-- Fields required for history management -->
    <form id="history-form" class="x-hidden">
    <input type="hidden" id="x-history-field"/>
    </form>
  </body>
</html>
 
Last edited:
Hi, sorry if it sounds dumb, but I'd like to be sure:

you have NOT any filename similar to /etc/apt/apt.conf.d/no-nag-script , right?

And only original repositories under /etc/apt ?
 
Last edited:
  • Like
Reactions: MarkusKo
One thing I have learnt in the 20 years I spent at the bleeding edge of tech is that there are no stupid Qs. I have seen the most remarkably able people miss the most basic things - so if you have anything you want to check, please ask away ...

Here are the outputs. The only thing I think is non-standard is rkhunter but that has been on my systems for a couple of years and is on all the other ones which are not suffering blank screens.

I am beginning to think the easiest thing might be to;
- remove the pve from the cluster,
- remove the data drives,
- reinstall pve,
- rejoin the cluster,
- reattach the zfs data drives

I have backed up everything to a pbs and can migrate the critical apps to the 2nd pve while I reinstall. Just need to check how to remove from and rejoin the cluster, and re-attach the data disks before I start.

Obv prefer to get it fixed if poss, but the above is a fallback I can live with.

Code: 
ls -a /etc/apt/apt.conf.d/
.  ..  01autoremove  10pveapthook  20listchanges  70debconf  76pveconf    76pveproxy  90rkhunter

Code: 
ls -a /etc/apt/
.   apt.conf.d     keyrings       listchanges.conf.d  sources.list    trusted.gpg.d
..  auth.conf.d  listchanges.conf  preferences.d       sources.list.d
 
Last edited:
You must log in or register to reply here.
Top Bottom