19:38:54 <swinokur> okay - so after changing the ESXI SCSI device from LSI SAS to LSI Parallel, I am now seeing this in dmesg:
19:38:54 <swinokur> Feb  4 12:51:27 baybee scsi: [ID 107833 kern.warning] WARNING: /pci@0,0/pci15ad,1976@10 (mpt0):
19:38:54 <swinokur> Feb  4 12:51:27 baybee  Disconnected command timeout for Target 0
19:38:54 <swinokur> Feb  4 12:51:27 baybee scsi: [ID 107833 kern.warning] WARNING: /pci@0,0/pci15ad,1976@10 (mpt0):
19:38:54 <swinokur> Feb  4 12:51:27 baybee  scsi_status=22, ioc_status=4b
19:38:55 <swinokur> Feb  4 12:51:27 baybee scsi: [ID 107833 kern.warning] WARNING: /pci@0,0/pci15ad,1976@10 (mpt0):
19:38:55 <swinokur> Feb  4 12:51:27 baybee  mpt_process_intr: invalid scsi status
19:39:33 <swinokur> 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? 
22:15:09 <danmcd> @sjorge --> you reported the big problem with BHYVE guests and large-MTU.  See https://smartos.org/bugview/OS-8620 for some encouraging results.
22:15:10 <fenix> → OS-8620: Early pull of illumos#13992 and illumos#17148 (Open)
22:17:52 <danmcd> fenix illumos#13992 is in the RTI queue.
22:17:53 <fenix> BUG 13992: VNIC in Zone loses access to properties (In Progress)
22:17:53 <fenix> ↳ https://www.illumos.org/issues/13992 | https://code.illumos.org/c/illumos-gate/+/4001
22:21:15 <danmcd> And it looks like pmooney is RTI-ing illumos#17148 very soon as well.
23:19:36 <pmooney> doing the last few checks on the built bits