okay - so after changing the ESXI SCSI device from LSI SAS to LSI Parallel, I am now seeing this in dmesg:

Feb 4 12:51:27 baybee scsi: [ID 107833 kern.warning] WARNING: /pci@0,0/pci15ad,1976@10 (mpt0):

Feb 4 12:51:27 baybee Disconnected command timeout for Target 0

Feb 4 12:51:27 baybee scsi: [ID 107833 kern.warning] WARNING: /pci@0,0/pci15ad,1976@10 (mpt0):

Feb 4 12:51:27 baybee scsi_status=22, ioc_status=4b

Feb 4 12:51:27 baybee scsi: [ID 107833 kern.warning] WARNING: /pci@0,0/pci15ad,1976@10 (mpt0):

Feb 4 12:51:27 baybee mpt_process_intr: invalid scsi status

I'll sure take a kernel warning over a system hang, but I wonder if this is any help in tracing what might be the root cause? Seems like it might be an interaction issue with ESXI?

@sjorge --> you reported the big problem with BHYVE guests and large-MTU. See smartos.org/bugview/OS-8620 for some encouraging results.

→ OS-8620: Early pull of illumos#13992 and illumos#17148 (Open)

fenix illumos#13992 is in the RTI queue.

BUG 13992: VNIC in Zone loses access to properties (In Progress)

↳ illumos.org/issues/13992 | code.illumos.org/c/illumos-gate/+/4001

And it looks like pmooney is RTI-ing illumos#17148 very soon as well.

doing the last few checks on the built bits