[SOLVED] Unraid server (6.9.x 6.10.x) crash and is unreachable after WOL

[terox]

Fixed. Not a bug, but the "CPU pinning" panel could induce into confusions for "non-sys guys". By parts:

 

The solution:

My CPU has 12 cores (6/6). I had isolate the CPU with id 0 and 6. My VM has pinned the rest of cores (1,2,3,5,7...). This configuration is completely wrong. The isolated CPUs "are out" of the task scheduler. By this misunderstood configuration I was achiving a crash may be because teorically non-isolated CPUs was using for VM (and containers). After a change of pinned CPUs from 0,6 to 1-5,7-11 and pin the same isolated cores to VM, the issue was fixed. I reserved the CPU 0,6 for dockers and Unraid processes.

 

Why I think that CPU Pinning induce to confusions:

There is nothing that visual in the panel that alert you that you have (or could you have) a potential issue of performance or configuration. To know some of this problems, you need to install a "Fix common Problems" to have some clue. I think that It should be integrated into the panel.

Thank you for help and clues.