sommerfeld: sorry for a response almost 12 hours later... I had a long shift at work, and my IRC bouncer died in the middle of it so i could only see your first message about project(5) and your example

but thank you very much for telling me one of the use cases behind it

before i went to work i set up my system to have a project for users in 'gid=2000(users)', and i set it with FSS to give the entire project 6 out of 8 CPU cores i believe

gonna paste the single line here:

group.users:2000:Shared project for standard users::users:project.max-lwps=(priv,5000,deny);project.max-tasks=(priv,5000,deny);project.max-physical-memory=(priv,21474836480,deny);project.pool=user-pool

i also added some other stuff too

Woodstock: have you seen 'vioscsi0: request stuck, resetting target' when you were doing your virtio changes?

i think we have all of them merged in, but someone running with nutanix as the host reported it, but then a full reset made it go away (which is strange)

jbk: never, at least not under normal circumstances. the mpt_sas bug did trigger nasty resets, though.

[illumos-gate] 18248 libzfs: uri_handlers are maximally confused -- Andy Fiddaman <illumos⊙fn>

Woodstock: hrm.. ok.. now that i've slept let me go through the commits more closely to see if we're missing any fixes.. the fact that killing and restarting the VM makes me wonder if maybe we were leaving the vioscsi device in some odd state that wasn't getting reset by a regular reboot...

[illumos-gate] 18405 Panic in smbios_open() when booted in VirtualBox -- Toomas Soome <tsoome⊙mc>