Any improvement on the NVMe driver for 2017 as I see ongoing demand
for NVMe devices in ESXi in pass-through mode for a virtualized Illumos based SAN/NAS

ex
https://forums.servethehome.com/index.php?threads/nvme-napp-it-omnios-pass-through-in-esxi-date.11043/
http://www.hardwareluxx.de/community/f101/zfs-stammtisch-570052-295.html#post25169817


Merry Christmas to everyone
Gea


Am 26.10.2016 um 17:20 schrieb Sebastien Roy:
On Wed, Oct 26, 2016 at 11:07 AM, Yuri Pankov <yuri.pan...@nexenta.com> wrote:
On Wed, 26 Oct 2016 11:04:07 -0400, Sebastien Roy wrote:
On Wed, Oct 26, 2016 at 9:47 AM, Hans Rosenfeld
<hans.rosenf...@nexenta.com> wrote:
So perhaps apix/pcmplusmp need to do some cleanup work when interrupts
are released so that ESXi understands what happens there. I don't really
know.

Disabling the use of MSI-X in nvme on ESXi seems to workaround that
problem. I've asked that question before: There are already a few other
drivers in illumos that disable MSI-X when running on ESXi. Should nvme
do the same, or would it better to disable MSI-X on ESXi completely?

We're working on interrupt/CPU scalability for vmxnet3s that relies on
MSI-X (which is already being used reliably by the current driver
which uses one per device). I would say that disabling MSI-X for the
entire platform would be a non-starter from our perspective.

I think this is more about the devices in PCI passthrough mode (and not the
"native" ones), seen a lot of issues with MSI-X enabled there.
I see, thanks for the clarification.

-Seb


--

Guenther Alka, Dipl.-Ing. (FH)
Leiter des Rechenzentrums
head of computer center

Tel 07171 602 627
Fax 07171 69259
guenther.a...@hfg-gmuend.de
http://rz.hfg-gmuend.de


-------------------------------------------
illumos-discuss
Archives: https://www.listbox.com/member/archive/182180/=now
RSS Feed: https://www.listbox.com/member/archive/rss/182180/21175430-2e6923be
Modify Your Subscription: 
https://www.listbox.com/member/?member_id=21175430&id_secret=21175430-6a77cda4
Powered by Listbox: http://www.listbox.com

Reply via email to