Proxmox Restarts after couple of hours.

[Bytales]

So i managed to set up my WIndows VM with GPU passtrough and direct SSD passtrough in Proxmox, all works fine, but after a couple of hours, the whole PC dies.

Now i dont know if the whole ProxMox server restarts or the computer dies, or whatever happens.
It seems somehow that maybe the NVME where the ProxMox is located doesnt get detected any more, because when booting the server again, it doesnt see it. But after some restarts it sees it.

Doesnt ProxMox create some logs ?
Where are these located and how can i acces them ?

These logs might hold a clue as to what is happening to the system. Perhaps i can find out what goes wrong and i can fix it.

 

[dcsapak]

for logs see 'journalctl' or the files in /var/log

also on the webinterface there is a 'syslog' tab for each node

 

[AlexLup]

After last upgrade this behaviour actually stopped, that and moving around my 10gb NICs in all 3 nodes, but I found NO logs at all.

Is there any kernel debugger that could capture any bluescreens for future crashes ?

 

[guletz]

Salutare @AlexLup,

You could make in your enviroment a centralised syslog and netconsole. Then you can send any logs on this remote syslog(including kernel crashes if the network is not affected by the crash). Plese see here:

https://www.cyberciti.biz/tips/linux-netconsole-log-management-tutorial.html

Bafta

 

[AlexLup]

Salutare @AlexLup,

You could make in your enviroment a centralised syslog and netconsole. Then you can send any logs on this remote syslog(including kernel crashes if the network is not affected by the crash). Plese see here:

https://www.cyberciti.biz/tips/linux-netconsole-log-management-tutorial.html

Bafta

Hi,
Thanks for this, however my proxmox cluster is at home and all I was really asking for was a "minidump" of sorts. If memory serves right this should be in standard debian edition but not in proxmox. Might be wrong tho!

Multsam,
A