Re: [openstack-dev] [nova] Discussions for DPDK support in OpenStack

2017-05-08 Thread TETSURO NAKAMURA 

Thank you for information !

So, you mean the situation is not changed from the reference thread[1] 
in qemu-devel ML 3 years ago, and the difficulties of ivshmem you 
mentioned is described in this thread. Am I right?


[1] "[Qemu-devel] Why I advise against using ivshmem"
https://lists.linuxfoundation.org/pipermail/virtualization/2014-June/026767.html

On 2017/05/08 9:09, Daniel P. Berrange wrote:

On Fri, Apr 28, 2017 at 09:38:38AM +0100, sfinu...@redhat.com wrote:

On Fri, 2017-04-28 at 13:23 +0900, TETSURO NAKAMURA wrote:

Hi Nova team,

I'm writing this e-mail because I'd like to have a discussion about
DPDK support at OpenStack Summit in Boston.

We have developed a dpdk-based patch panel named SPP[1], and we'd
like to start working on Openstack (ML2 driver) to develop
"networking-spp".

Especially, we'd like to use DPDK-ivshmem that was used to be used
to create "dpdkr" interface in ovs-dpdk[2].


To the best of my knowledge, IVSHMEM ports are no longer supported in
upstream. The documentation for this feature was recently removed from
OVS [1] stating:

  - The ivshmem library has been removed in DPDK since DPDK 16.11.
  - The instructions/scheme provided will not work with current
supported and future DPDK versions.
  - The linked patch needed to enable support in QEMU has never
been upstreamed and does not apply to the last 4 QEMU releases.
  - Userspace vhost has become the defacto OVS-DPDK path to the guest.

Note: I worked on DPDK vSwitch [2] way back when, and there were severe
security implications with sharing a chunk of host memory between
multiple guests (which is how IVSHMEM works). I'm not at all surprised
the feature was killed.


Security is only one of the issues. Upstream QEMU maintainers considered
the ivshmem device to have a seriously flawed design and discourage anyone
from using it. For anything network related QEMU maintainers strongly
recommand using vhost-user.

IIUC, there is some experimental work to create a virtio based replacement
for ivshmem, for non-network related vm-2-vm communications, but that is
not going to be something usable for a while yet. This however just
reinforces the point that ivshmem is considered obsolete / flawed
technology by QEMU maintainers.

Regards,
Daniel



--
Tetsuro Nakamura 
NTT Network Service Systems Laboratories
TEL:0422 59 6914(National)/+81 422 59 6914(International)
3-9-11, Midori-Cho Musashino-Shi, Tokyo 180-8585 Japan



__
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev

Re: [openstack-dev] [nova] Discussions for DPDK support in OpenStack

2017-05-08 Thread Daniel P. Berrange 
On Fri, Apr 28, 2017 at 09:38:38AM +0100, sfinu...@redhat.com wrote:
> On Fri, 2017-04-28 at 13:23 +0900, TETSURO NAKAMURA wrote:
> > Hi Nova team,
> > 
> > I'm writing this e-mail because I'd like to have a discussion about
> > DPDK support at OpenStack Summit in Boston.
> > 
> > We have developed a dpdk-based patch panel named SPP[1], and we'd
> > like to start working on Openstack (ML2 driver) to develop
> > "networking-spp".
> > 
> > Especially, we'd like to use DPDK-ivshmem that was used to be used
> > to create "dpdkr" interface in ovs-dpdk[2].
> 
> To the best of my knowledge, IVSHMEM ports are no longer supported in
> upstream. The documentation for this feature was recently removed from
> OVS [1] stating:
> 
>   - The ivshmem library has been removed in DPDK since DPDK 16.11.
>   - The instructions/scheme provided will not work with current
>     supported and future DPDK versions.
>   - The linked patch needed to enable support in QEMU has never
>     been upstreamed and does not apply to the last 4 QEMU releases.
>   - Userspace vhost has become the defacto OVS-DPDK path to the guest.
> 
> Note: I worked on DPDK vSwitch [2] way back when, and there were severe
> security implications with sharing a chunk of host memory between
> multiple guests (which is how IVSHMEM works). I'm not at all surprised
> the feature was killed.

Security is only one of the issues. Upstream QEMU maintainers considered
the ivshmem device to have a seriously flawed design and discourage anyone
from using it. For anything network related QEMU maintainers strongly
recommand using vhost-user.

IIUC, there is some experimental work to create a virtio based replacement
for ivshmem, for non-network related vm-2-vm communications, but that is
not going to be something usable for a while yet. This however just
reinforces the point that ivshmem is considered obsolete / flawed
technology by QEMU maintainers.

Regards,
Daniel
-- 
|: https://berrange.com  -o-https://www.flickr.com/photos/dberrange :|
|: https://libvirt.org -o-https://fstop138.berrange.com :|
|: https://entangle-photo.org-o-https://www.instagram.com/dberrange :|

__
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev

Re: [openstack-dev] [nova] Discussions for DPDK support in OpenStack

2017-05-08 Thread TETSURO NAKAMURA 

Thank you for reply!

On 2017/04/28 4:38, sfinu...@redhat.com wrote:

On Fri, 2017-04-28 at 13:23 +0900, TETSURO NAKAMURA wrote:

Hi Nova team,

I'm writing this e-mail because I'd like to have a discussion about
DPDK support at OpenStack Summit in Boston.

We have developed a dpdk-based patch panel named SPP[1], and we'd
like to start working on Openstack (ML2 driver) to develop
"networking-spp".

Especially, we'd like to use DPDK-ivshmem that was used to be used
to create "dpdkr" interface in ovs-dpdk[2].


To the best of my knowledge, IVSHMEM ports are no longer supported in
upstream. The documentation for this feature was recently removed from
OVS [1] stating:

  - The ivshmem library has been removed in DPDK since DPDK 16.11.
  - The instructions/scheme provided will not work with current
supported and future DPDK versions.
  - The linked patch needed to enable support in QEMU has never
been upstreamed and does not apply to the last 4 QEMU releases.
  - Userspace vhost has become the defacto OVS-DPDK path to the guest.

Note: I worked on DPDK vSwitch [2] way back when, and there were severe
security implications with sharing a chunk of host memory between
multiple guests (which is how IVSHMEM works). I'm not at all surprised
the feature was killed.


We have issued a blueprint[3] for that use case.


Per above, I don't think this is necessary. vhost-user ports already
work as expected in nova.



Yes, IVSHMEM is a critical issue for multitenancy.
Still, we'd like to state that there is a use case for private cloud 
such as carrier NFV in which sharing host memory doesn't become a 
critical issue. In that use case we'd like to use ivshmem for good 
performance.



As we are attending Boston Summit, could you have a discussion with
us at the Summit?


I'll be around the summit (IRC: sfinucan) if you want to chat more.
However, I'd suggest reaching out to Sean Mooney or Igor Duarte Cardoso
(both CCd) if you want further information about general support of
OVS-DPDK in OpenStack and DPDK acceleration in SFC, respectively. I'd
also suggest looking at networking-ovs-dpdk [3] which contains a lot of
helper tools for using OVS-DPDK in OpenStack, along with links to a
Brighttalk video I recently gave regarding the state of OVS-DPDK in
OpenStack.


Thank you very much for information !
I'm already reaching out to Sean Mooney in another case,
so I try to reach out to Igor Duarte Cardoso-san.



Hope this helps,
Stephen

[1] https://github.com/openvswitch/ovs/commit/90ca71dd317fea1ccf0040389
dae895aa7b2b561
[2] https://github.com/01org/dpdk-ovs
[3] https://github.com/openstack/networking-ovs-dpdk

__
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev



--
Tetsuro Nakamura 
NTT Network Service Systems Laboratories
TEL:0422 59 6914(National)/+81 422 59 6914(International)
3-9-11, Midori-Cho Musashino-Shi, Tokyo 180-8585 Japan



__
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev

Re: [openstack-dev] [nova] Discussions for DPDK support in OpenStack

2017-04-28 Thread sfinucan 
On Fri, 2017-04-28 at 13:23 +0900, TETSURO NAKAMURA wrote:
> Hi Nova team,
> 
> I'm writing this e-mail because I'd like to have a discussion about
> DPDK support at OpenStack Summit in Boston.
> 
> We have developed a dpdk-based patch panel named SPP[1], and we'd
> like to start working on Openstack (ML2 driver) to develop
> "networking-spp".
> 
> Especially, we'd like to use DPDK-ivshmem that was used to be used
> to create "dpdkr" interface in ovs-dpdk[2].

To the best of my knowledge, IVSHMEM ports are no longer supported in
upstream. The documentation for this feature was recently removed from
OVS [1] stating:

  - The ivshmem library has been removed in DPDK since DPDK 16.11.
  - The instructions/scheme provided will not work with current
    supported and future DPDK versions.
  - The linked patch needed to enable support in QEMU has never
    been upstreamed and does not apply to the last 4 QEMU releases.
  - Userspace vhost has become the defacto OVS-DPDK path to the guest.

Note: I worked on DPDK vSwitch [2] way back when, and there were severe
security implications with sharing a chunk of host memory between
multiple guests (which is how IVSHMEM works). I'm not at all surprised
the feature was killed.

> We have issued a blueprint[3] for that use case.

Per above, I don't think this is necessary. vhost-user ports already
work as expected in nova.

> As we are attending Boston Summit, could you have a discussion with
> us at the Summit?

I'll be around the summit (IRC: sfinucan) if you want to chat more.
However, I'd suggest reaching out to Sean Mooney or Igor Duarte Cardoso
(both CCd) if you want further information about general support of
OVS-DPDK in OpenStack and DPDK acceleration in SFC, respectively. I'd
also suggest looking at networking-ovs-dpdk [3] which contains a lot of
helper tools for using OVS-DPDK in OpenStack, along with links to a
Brighttalk video I recently gave regarding the state of OVS-DPDK in
OpenStack.

Hope this helps,
Stephen

[1] https://github.com/openvswitch/ovs/commit/90ca71dd317fea1ccf0040389
dae895aa7b2b561
[2] https://github.com/01org/dpdk-ovs
[3] https://github.com/openstack/networking-ovs-dpdk

__
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev

Re: [openstack-dev] [nova] Discussions for DPDK support in OpenStack

2017-04-28 Thread Guo, Ruijing 
Get information that  DPDK removed ivshmem
from http://dpdk.org/ml/archives/dev/2016-July/044552.html

-Original Message-
From: TETSURO NAKAMURA [mailto:nakamura.tets...@lab.ntt.co.jp] 
Sent: Friday, April 28, 2017 12:23 PM
To: openstack-dev@lists.openstack.org
Subject: [openstack-dev] [nova] Discussions for DPDK support in OpenStack

Hi Nova team,

I'm writing this e-mail because I'd like to have a discussion about DPDK 
support at OpenStack Summit in Boston.

We have developed a dpdk-based patch panel named SPP[1], and we'd like to start 
working on Openstack (ML2 driver) to develop "networking-spp".

Especially, we'd like to use DPDK-ivshmem that was used to be used to create 
"dpdkr" interface in ovs-dpdk[2].

We have issued a blueprint[3] for that use case.

As we are attending Boston Summit, could you have a discussion with us at the 
Summit?

[1] http://www.dpdk.org/browse/apps/spp/
[2]
http://openvswitch.org/support/dist-docs-2.5/INSTALL.DPDK.md.html#L446-L490
[3] https://blueprints.launchpad.net/nova/+spec/libvirt-options-for-dpdk

Sincerely,

--
Tetsuro Nakamura <nakamura.tets...@lab.ntt.co.jp> NTT Network Service Systems 
Laboratories
TEL:0422 59 6914(National)/+81 422 59 6914(International) 3-9-11, Midori-Cho 
Musashino-Shi, Tokyo 180-8585 Japan



__
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev
__
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev

[openstack-dev] [nova] Discussions for DPDK support in OpenStack

2017-04-27 Thread TETSURO NAKAMURA 

Hi Nova team,

I'm writing this e-mail because I'd like to have a discussion about DPDK  
support at OpenStack Summit in Boston.


We have developed a dpdk-based patch panel named SPP[1], and we'd like  
to start working on Openstack (ML2 driver) to develop "networking-spp".


Especially, we'd like to use DPDK-ivshmem that was used to be used to  
create "dpdkr" interface in ovs-dpdk[2].


We have issued a blueprint[3] for that use case.

As we are attending Boston Summit, could you have a discussion with us  
at the Summit?


[1] http://www.dpdk.org/browse/apps/spp/
[2]  
http://openvswitch.org/support/dist-docs-2.5/INSTALL.DPDK.md.html#L446-L490

[3] https://blueprints.launchpad.net/nova/+spec/libvirt-options-for-dpdk

Sincerely,

--
Tetsuro Nakamura 
NTT Network Service Systems Laboratories
TEL:0422 59 6914(National)/+81 422 59 6914(International)
3-9-11, Midori-Cho Musashino-Shi, Tokyo 180-8585 Japan



__
OpenStack Development Mailing List (not for usage questions)
Unsubscribe: openstack-dev-requ...@lists.openstack.org?subject:unsubscribe
http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev