PVE Stuck in Degraded State

[dc-ko]

Hello,
I have a 4 node proxmox cluster in which two hosts experienced a power outage. Everything returns fine in Proxmox, however Pulse shows them in a degraded state. Unsure how to clear the degraded flag.

Yellow Dots on main page:

Cluster Status Degraded:

Test of Connectivity Shows Green:

Alerts show status resolved:

[rcourtman]

Looking at the screenshots, the connectivity test shows green and the cluster is reachable through pve-1, but the individual node errors for pve-3 and pve-4 persist showing timeouts to their direct IPs (192.168.2.44 and 192.168.2.169).

A few questions:

1. Can Pulse (running on pve-1) directly reach pve-3 and pve-4 on their management interfaces? From the Pulse LXC, try:

   curl -k https://192.168.2.44:8006/api2/json/version
   curl -k https://192.168.2.169:8006/api2/json/version

   If these timeout, there might be a firewall or network issue blocking direct node access.

2. For now, you can try restarting Pulse to clear the cached error state:

   systemctl restart pulse

   The cluster client should re-initialize and probe the nodes fresh.

If the curl commands fail but Proxmox itself sees the nodes as online, this suggests network segmentation—Pulse can reach pve-1 but not the other nodes directly. In that case, Pulse will continue showing degraded until the network path is fixed.

If the curl commands succeed but Pulse still shows degraded after restart, let me know and I will dig deeper into the recovery logic.

[dc-ko]

Thanks a lot for looking. The curl command succeeded as far as reachable - hit 401 on all 4 nodes. Pulse running on pve-2; it was installed originally on pve-1.
I ran a systemctl restart pulse and was greeted with Pulse reachable on all 4 nodes right after.

curl test:

After pulse restart:

[dc-ko]

Perhaps related to #754 and resolved already? I am on 4.35.0