Since it's possible to delete memory pool in DPDK
we can try to estimate better required memory size
when port is reconfigured, e.g. with different number
of rx queues.
Signed-off-by: Robert Wojciechowicz <robertx.wojciechow...@intel.com>
Acked-by: Ian Stokes <ian.sto...@intel.co
On Wed, Apr 05, 2017 at 05:12:44PM +0100, Kevin Traynor wrote:
> On 02/24/2017 11:27 AM, Robert Wojciechowicz wrote:
> > Since it's possible to delete memory pool in DPDK
> > we can try to estimate better required memory size
> > when port is reconfigured, e.g. with diffe
On Thu, Mar 09, 2017 at 01:59:51PM +, Robert Wojciechowicz wrote:
> When using vhost-user client or server mode with openstack,
> neutron needs to be able to construct the fully qualified socket path
> and pass it to nova.
> While the relative vhost-user socket directory is
On Fri, Feb 24, 2017 at 11:27:12AM +, Robert Wojciechowicz wrote:
> Since it's possible to delete memory pool in DPDK
> we can try to estimate better required memory size
> when port is reconfigured, e.g. with different number
> of rx queues.
>
> Signed-off-by: Ro
On Wed, Mar 08, 2017 at 09:31:22PM -0800, Ben Pfaff wrote:
> On Thu, Mar 02, 2017 at 02:40:19PM +0000, Robert Wojciechowicz wrote:
> > When using vhost-user client or server mode with openstack,
> > neutron needs to be able to construct the fully qualified socket path
> >
that integrate
with Open vSwitch it is a perfect place for exposing OVS
run directory.
So this patch adds to the `Open_vSwitch` table `external_ids` column
this key:
1. ovs-run-dir- OVS run directory
Signed-off-by: Robert Wojciechowicz <robertx.wojciechow...@intel.com>
Acked-by: Sean K
Expose existing netdev stats via sFlow.
Export sFlow ETHERNET structure with available counters.
Map existing stats to counters in the GENERIC INTERFACE
sFlow structure.
Adjust unit test to accommodate these new counters.
Signed-off-by: Robert Wojciechowicz <robertx.wojciechow...@intel.com>
Since it's possible to delete memory pool in DPDK
we can try to estimate better required memory size
when port is reconfigured, e.g. with different number
of rx queues.
Signed-off-by: Robert Wojciechowicz <robertx.wojciechow...@intel.com>
Acked-by: Ian Stokes <ian.sto...@intel.co
Since it's possible to delete memory pool in DPDK
we can try to estimate better required memory size
when port is reconfigured, e.g. with different number
of rx queues.
Signed-off-by: Robert Wojciechowicz <robertx.wojciechow...@intel.com>
Acked-by: Ian Stokes <ian.sto...@intel.com&
On Fri, Feb 03, 2017 at 08:53:32AM +, Stokes, Ian wrote:
>
> Looks OK to me, tested without issue.
>
> Acked-by: Ian Stokes
Many thanks Ian for testing this patch.
Br,
Robert
___
dev mailing list
d...@openvswitch.org
On Thu, Feb 02, 2017 at 10:00:12AM -0800, Neil McKee wrote:
> Robert,
>
> It's great to get the multicast and broadcast counters in there.
>
> I do wonder if it's worth including the whole ethernet stats block just to
> distinguish alignment / FCS / frame-too-long errors, given that total
>
with having 64 queues for dpdk physical ports
> which causes ovs to crash.
Thanks for testing
>
> A few other comments below.
> >
> > Signed-off-by: Robert Wojciechowicz <robertx.wojciechow...@intel.com>
> > ---
> > lib/netdev-dpdk.c | 85 ++
On Wed, Nov 09, 2016 at 08:02:49AM -0500, Robert Wojciechowicz wrote:
>
> Hi,
>
> I started working on this patch after this discussion:
> https://mail.openvswitch.org/pipermail/ovs-discuss/2016-September/042643.html
>
> It turned out that it is not possible in OVS to
Expose existing netdev stats via sFlow.
Export sFlow ETHERNET structure with available counters.
Map existing stats to counters in the GENERIC INTERFACE
sFlow structure.
Adjust unit test to accommodate these new counters.
Signed-off-by: Robert Wojciechowicz <robertx.wojciechow...@intel.
On Fri, Nov 18, 2016 at 12:35:13PM -0700, Neil McKee wrote:
> That sounds like a robust way to do it, yes. Does it have the desired
> effect? Does it eliminate this structure for all the virtual ethernet
> ports? I believe there should only be <= 1 physical ethernet port on each
It
On Thu, Nov 17, 2016 at 10:21:23AM -0800, Neil McKee wrote:
> I'm wondering if the ethernet stats block is going to mean anything for
> virtual interfaces. What if we only export it for physical interfaces?
> That might reduce the overhead significantly. (Maybe your patch already
> does this -
On Tue, Nov 15, 2016 at 11:13:10PM -0700, Neil McKee wrote:
> Hi Robert,
>
> (3) adding vlan counters for every vlan that an interface belongs to -- I
> not sure about this one. What if an interface is on a whole set of vlans?
> This vlan counters structure has hardly ever been used.
Expose existing netdev stats via sFlow.
Export sFlow ETHERNET structure with available counters.
Map existing stats to counters in the GENERIC INTERFACE
sFlow structure.
Export sFlow VLAN structure with available counters.
Adjust unit test to accommodate these new counters.
Signed-off-by: Robert
18 matches
Mail list logo