Stux
MVP
- Joined
- Jun 2, 2016
- Messages
- 4,419
Just performed a migration to SCALE 23.10.2 from one of my CORE servers.
I ran two VMs on CORE, one for pfSense with NIC PCIe passthrough, and the other an Ubuntu docker machine for running docker compose apps. I moved away from Jails years ago now, and am looking forward to jailmaker in Dragonfish.
The migration process is fairly impressive. Everything went fairly smoothly, until I had trouble bring up an NFS mount due to missing user/group issue
I had a pfSense bhyve VM, with two NICs pci passed through, amazingly this migrated almost without issue, just had to reselect the NIC devices at their new location
Perhaps this could've been migrated too ;)
Only other thing I had to do was set the CPU Mode to Host Model (or Passthrough) so that pfSense would see the AES-NI capability. It would've been helpful if CPU Mode had a tool tip on it. In fact... it would be helpful if all fields had tool tips, but at least they can be found in the docs.
Then my ubuntu VM. In TrueNAS, it was a simple matter to share the same NIC with all VMs and TrueNAS, and as everyone knows by now (except me at the time), you need to setup a bridge to do the same thing on SCALE.... and the Accessing NAS from VM Tutorial is gold for this.
Once I had that done, then it was just a matter of updating the netplan in the VM, as the network device had been renamed... its now
And... then solving the NFS issue.
BUT, everything seems pretty good... was super impressed with how quickly I managed to get my pfSense VM back up, altho I'm still a bit sour about blowing 8 hours on a frickin checkmark.
And SCALE is looking pretty good.
AND hopefully my VMs won't randomly lock up due to the BSD/bhyve bug now
I ran two VMs on CORE, one for pfSense with NIC PCIe passthrough, and the other an Ubuntu docker machine for running docker compose apps. I moved away from Jails years ago now, and am looking forward to jailmaker in Dragonfish.
The migration process is fairly impressive. Everything went fairly smoothly, until I had trouble bring up an NFS mount due to missing user/group issue
I had a pfSense bhyve VM, with two NICs pci passed through, amazingly this migrated almost without issue, just had to reselect the NIC devices at their new location
5/0/0
and 5/0/1
became 0000:05:00.0
and 0000:05:00.1
Perhaps this could've been migrated too ;)
Only other thing I had to do was set the CPU Mode to Host Model (or Passthrough) so that pfSense would see the AES-NI capability. It would've been helpful if CPU Mode had a tool tip on it. In fact... it would be helpful if all fields had tool tips, but at least they can be found in the docs.
Then my ubuntu VM. In TrueNAS, it was a simple matter to share the same NIC with all VMs and TrueNAS, and as everyone knows by now (except me at the time), you need to setup a bridge to do the same thing on SCALE.... and the Accessing NAS from VM Tutorial is gold for this.
Once I had that done, then it was just a matter of updating the netplan in the VM, as the network device had been renamed... its now
ens3
instead of ens4p1
or whatever it was before.And... then solving the NFS issue.
BUT, everything seems pretty good... was super impressed with how quickly I managed to get my pfSense VM back up, altho I'm still a bit sour about blowing 8 hours on a frickin checkmark.
And SCALE is looking pretty good.
AND hopefully my VMs won't randomly lock up due to the BSD/bhyve bug now