churchers / vm-bhyve

Shell based, minimal dependency bhyve manager
BSD 2-Clause "Simplified" License
839 stars 181 forks source link

Host crashes on all installs #500

Closed garrettboone closed 1 year ago

garrettboone commented 1 year ago

I am brand new to bhyve but have already established a FreeBSD 13.1 host and can do jails and linux compat without any problems. I wanted to run a Debian or Ubuntu VM, and am getting stuck.

I was trying to run bhyve in a jail. I have also tried running it on the host, but the follow results are the same either way (which I think is expected, although I don't know what the problem is):

I can get a grub loader to show, but can't get past it. I experimented with different settings, tried uefi, etc. I've tried with install -f and without. But, I get a blank screen. I saw the issue about setting screen resolution, though ultimately, even when I try to follow the testvm example with a FreeBSD 13.1 iso, it causes the host to crash before I can even finish typing vm console testvm.

Any idea what might be causing this?

From the testvm log:

Jan 30 03:18:11: generated static mac 58:9c:fc:06:c0:db (based on 'testvm:0:1675048691:0')
Jan 30 03:22:05: initialising
Jan 30 03:22:05:  [loader: bhyveload]
Jan 30 03:22:05:  [cpu: 1]
Jan 30 03:22:05:  [memory: 256M]
Jan 30 03:22:05:  [hostbridge: standard]
Jan 30 03:22:05:  [com ports: com1]
Jan 30 03:22:05:  [uuid: b9b589c9-a04c-11ed-93a0-589cfc006070]
Jan 30 03:22:05:  [debug mode: no]
Jan 30 03:22:05:  [primary disk: disk0.img]
Jan 30 03:22:05:  [primary disk dev: file]
Jan 30 03:22:05: booting
Jan 30 03:22:05: bhyveload -c /dev/nmdm-testvm.1A -m 256M -e autoboot_delay=3 -d /usr/local/vms/.iso/FreeBSD-13.1-RELEASE-amd64-disc1.iso testvm

From the last Debian attempt:

Jan 30 03:07:47: generated static mac 58:9c:fc:0c:21:b0 (based on 'vmDebian116:0:1675048067:0')
Jan 30 03:08:00: initialising
Jan 30 03:08:00:  [loader: grub]
Jan 30 03:08:00:  [cpu: 1]
Jan 30 03:08:00:  [memory: 512M]
Jan 30 03:08:00:  [hostbridge: standard]
Jan 30 03:08:00:  [com ports: com1]
Jan 30 03:08:00:  [uuid: 4614419b-a04b-11ed-8275-589cfc006070]
Jan 30 03:08:00:  [debug mode: no]
Jan 30 03:08:00:  [primary disk: disk0.img]
Jan 30 03:08:00:  [primary disk dev: file]
Jan 30 03:08:00: booting
Jan 30 03:08:00: create file /usr/local/vms/vmDebian116/device.map
Jan 30 03:08:00:  -> (cd0) /usr/local/vms/.iso/debian-11.6.0-amd64-netinst.iso
Jan 30 03:08:00:  -> (hd0) /usr/local/vms/vmDebian116/disk0.img
Jan 30 03:08:00: /usr/local/sbin/grub-bhyve -c /dev/nmdm-vmDebian116.1A -m /usr/local/vms/vmDebian116/device.map -M 512M -r cd0 vmDebian116

I am running a droplet at Digital Ocean, 8GB RAM, 4vcpu. I have looked through system logs and jail logs but cannot find anything else. I suspect it has something to do with the droplet.

# dmesg | grep Features2
Features2=0xfffa3223<SSE3,PCLMULQDQ,VMX,SSSE3,FMA,CX16,PCID,SSE4.1,SSE4.2,x2APIC,MOVBE,POPCNT,TSCDLT,AESNI,XSAVE,OSXSAVE,AVX,F16C,RDRAND,HV>AMD Features2=0x21<LAHF,ABM>
garrettboone commented 1 year ago

I got a bit further by adding serial_boot="YES" and hititng 3 on testvm.

Type '?' for a list of commands, 'help' for more detailed help.
OK set boot_serial=NO
OK boot
Loading kernel...
/boot/kernel/kernel text=0x184d70 text=0xdfdfc0 text=0x6634c4 data=0x140 data=0x1be3b8+0x440c48 syms=[0x8+0x188d78+0x8+0x1a7803]
Loading configured modules...
can't find '/etc/hostid'
can't find '/boot/entropy'

But then droplet hangs again.

garrettboone commented 1 year ago

Couldn't figure it out, closing.