Nodes keep shutting down at boot time.

[Belokan]

Hello,

I have a 4.2.5 PVE cluster that was running fine for few weeks but now I have an issue I can't figure out.

This is a basic cluster based on 2 physical nodes and one VM running on a NAS. I have a single vlan and storage is based on 2 NAS (nfs/iscsi). I've created an HA group with the 2 physical nodes and created a few resources (debian/centos VMs).

"Something" went wrong this morning and I've lost connectivity from office to home. When I came back, both physical nodes and the VM were powered off.

Now each time I try to power on the nodes, they start and then shutdown themselves few seconds after.
It looks like it's possible to keep them "up" if I unplug the network, but as soon as I plug it again it shutdown right after.

As the 2 physical nodes have no screen/keyboard, I've made my tests to solve the problem on the VM's console ...

Any idea ?

Thanks in advance !

 

[Belokan]

Forget about it, for some reason it's the nut/upsmon client that went crazy... So nodes were shutting down at boot because upsmon failed to connect to the server on both NAS.

I'll have to investigate this deeper... Should you mark the thread as solved or just delete it as it does not involve PM at all ?