00:11:17 <tozhu> hello all, can we startup a VM under (or bind with) a NUMA node? I mean add some code into ‘vmadm start’ 
19:05:32 <pjustice> >>> vminfod watchForEvent "VM.js receive global (ac748f8e-a908-e8dc-c388-84acefb21ede)" timeout exceeded
19:05:38 <pjustice> vmadm send/receive
19:05:59 <pjustice> Any thoughts about what I might try to patch up to get this to work? repeatable.
19:06:13 <pjustice> The ZFS datasets get transferred, apparently.
19:06:24 <pjustice> base image exists
19:14:12 <pjustice> clock time a little under 1 minute 30
19:14:35 <pjustice> source zone is stopped, set not to autoboot
19:14:38 <pjustice> 4 GB RAM
19:15:05 <pjustice> zfs dataset is small, a bit over 5 GB
21:04:10 <danmcd_> ptree of receiving process, and pstacks of relevant ones, might help.  
21:04:25 <danmcd_> Also what VM type are you sending?  If it's BHYVE, there are relatively recent fixes.
21:05:03 <danmcd_> Look for this comment:
21:05:04 <danmcd_>             // This event stream is used globally while vmbundle is forked many 
21:05:28 <danmcd_> in VM.js as that's what's being fired and getting flagged as timeout-exceeded by vminfod.
21:05:42 <danmcd> fixed nick
21:13:20 <pjustice> os zone
21:22:39 <pjustice> https://yagi.h-net.org/vminfod_timeout_treestack.txt
21:22:53 <pjustice> Not sure how useful that is, if I don't get it just as it's about to croak
21:27:58 <danmcd> Yeah... need to just-before-croak version.  (SOrry gotta run on an errand...)
21:31:11 <pjustice> No worries, it _is_ Friday night.