CMD access is good, GUI access is bad

[BobParr222]

PVE version is 7.1
If I go direct to my host and login via the command line, I have no issues getting authenticated and logged in, and can do all things accessible through the command line. However, when accessing the GUI, I get "Login failed. Please try again" and cannot progress beyond to manage my VMs. I have verified I can, from the command line, successfully ping other assets on the network, as well as assets on the internet. From my laptop I can ping the host, and pull up the GUI interface, but it will not let me login. Where do I start my troubleshooting?

~Bob

 

[datschlatscher]

Are you trying to log in as root or as a regular user?

By default, Proxmox assumes the `root` user is the system administrator. Login with root is always possible.
You will not be able to log in as a regular user right away, even if they have sudo privileges. You can allow more users to log into Proxmox by configuring authentication realms or setting up something like LDAP.

 

[BobParr222]

Are you trying to log in as root or as a regular user?

By default, Proxmox assumes the `root` user is the system administrator. Login with root is always possible.
You will not be able to log in as a regular user right away, even if they have sudo privileges. You can allow more users to log into Proxmox by configuring authentication realms or setting up something like LDAP.

Root. I use the same creds when logging in via command line as I do the GUI, but the GUI fails each time, whereas the command line works every time.

 

[dcsapak]

what does the journal/syslog show at the time the login fails?

 

[BobParr222]

what does the journal/syslog show at the time the login fails?

root@PROX1:~# tail /var/log/syslog
Mar 15 07:35:11 PROX1 pvescheduler[84349]: jobs: cfs-lock 'file-jobs_cfg' error: no quorum!
Mar 15 07:35:11 PROX1 pvescheduler[84348]: replication: cfs-lock 'file-replication_cfg' error: no quorum!
Mar 15 07:36:11 PROX1 pvescheduler[84482]: jobs: cfs-lock 'file-jobs_cfg' error: no quorum!
Mar 15 07:36:11 PROX1 pvescheduler[84481]: replication: cfs-lock 'file-replication_cfg' error: no quorum!
Mar 15 07:37:11 PROX1 pvescheduler[84616]: jobs: cfs-lock 'file-jobs_cfg' error: no quorum!
Mar 15 07:37:11 PROX1 pvescheduler[84615]: replication: cfs-lock 'file-replication_cfg' error: no quorum!
Mar 15 07:38:11 PROX1 pvescheduler[84752]: jobs: cfs-lock 'file-jobs_cfg' error: no quorum!
Mar 15 07:38:11 PROX1 pvescheduler[84751]: replication: cfs-lock 'file-replication_cfg' error: no quorum!
Mar 15 07:38:19 PROX1 pvedaemon[966]: Cluster not quorate - extending auth key lifetime!
Mar 15 07:38:28 PROX1 pvedaemon[966]: authentication failure; rhost=::ffff:172.22.22.103 user=root@pam msg=cfs-lock 'file-priv_tfa_cfg' error: no quorum!
root@PROX1:~#


I'm currently reading up on corosync.conf (in relation to the 'no quorum!' error - not sure if this is a separate issue?), and it appears that one should be able to edit that file (per the wiki: https://pve.proxmox.com/pve-docs/chapter-pvecm.html#_corosync_configuration). However, I get "Permission denied" when making the attempt.

root@PROX1:~# cp /etc/pve/corosync.conf /etc/pve/corosync.conf.new
cp: cannot create regular file '/etc/pve/corosync.conf.new': Permission denied
root@PROX1:~#

 

[dcsapak]

it seems your cluster is not quorate, that can have some side effects (like the login not working)
when you don't have quorum /etc/pve is read-only since the changes cannot be distributed across the cluster

i'd try to find out *why* it's not quorate and fix the underlying issue

 

[BobParr222]

it seems your cluster is not quorate, that can have some side effects (like the login not working)
when you don't have quorum /etc/pve is read-only since the changes cannot be distributed across the cluster

i'd try to find out *why* it's not quorate and fix the underlying issue

Thanks! I will reply once I have that determined.

 

[BobParr222]

@dcsapak Thanks again! I broke up the cluster (https://pve.proxmox.com/pve-docs/chapter-pvecm.html#_remove_a_cluster_node) into their respective individual nodes/hosts, and can now log in via GUI again.

The *why* was a forced network change, which included IP schema change, which adversely affected the cluster/quorum. I should be able to get the cluster back together, and the quorum should then be resolved also.