Open cuinutanix opened 7 years ago
I did about 800 cycles of destroy/start on a VM backed with hugepages (4 cores x 8 threads, 8GB of memory) on a P8 machine here without seeing any lockups or any misbehaviour at all. Clearly there is something more needed to trigger this bug.
Seeing this multiple times per day while doing manual power on and off.
Host is 2 sockets x 8 cores, 512GB memory. The VM being powered on is configured with 4 cores x 8 threads, with 8GB of memory, backed by hugetlbfs.
Host is running hostos-2.0 release from 2017-03-31. Some googling turned up this patch: https://lkml.org/lkml/2017/2/28/378.