bitwise_rscan() is found to be hot spot in ovn-controller during OVN
scalability tests. It is triggered by lflow_run() when processing
lflow updates from SB ovsdb. The perf result shows:
+ 35.90% ovn-controller ovn-controller [.] bitwise_rscan
+ 13.39% ovn-controller [kernel.kallsyms]
Adding ovs dev
On 3/15/16, 4:28 PM, "Darrell Ball" wrote:
>Signed-off-by: Darrell Ball
>---
> ovn/utilities/ovn-sbctl.c | 295 +-
> 1 file changed, 291 insertions(+), 4 deletions(-)
>
>diff --git a/ovn/utilities/ovn-sbctl.c b/ovn/utilities/ovn-sbct
Original Message
> From: Ben Warren
>
> Signed-off-by: Ben Warren
> ---
> include/openvswitch/automake.mk | 1 +
> include/openvswitch/geneve.h| 63 +++
++
> lib/automake.mk | 1 -
> lib/geneve.h| 63
---
On 15 March 2016 at 07:13, Ryan Moats wrote:
>
> Original Message
> > A (physical port + vlan) or (physical port + no tag) can be bound to a
> > logical switch. When one unbinds (physical port + no tag) from a logical
> > switch, the emulator inadvertantly deletes flows for (physical po
Hi Jesse,
Please find my answers inline.
Regards
_Sugesh
> -Original Message-
> From: Jesse Gross [mailto:je...@kernel.org]
> Sent: Thursday, March 17, 2016 11:50 PM
> To: Chandran, Sugesh
> Cc: dev@openvswitch.org
> Subject: Re: [ovs-dev] [RFC PATCH] tunneling: Improving vxlan performa
Currently we treat set_multiq() calls specially in netdev and dpif-netdev:
every pmd thread must be stopped and set_multiq() is allowed to destroy and
recreate the device.
I think we can improve this by:
* Generalizing the mechanism to allow changing other parameters at runtime
(such as MTU).
*
On Wed, Mar 16, 2016 at 8:15 PM, Ben Pfaff wrote:
> On Wed, Mar 16, 2016 at 07:16:46PM -0700, Russell Bryant wrote:
> > This was a mistake. ovs-testcontroller was duplicated in the ovn
> > package. It's already in the main openvswitch package, where it
> > belongs.
> >
> > Signed-off-by: Russel
On Thu, Mar 17, 2016 at 11:20:41AM +0200, Liran Schour wrote:
> > diff --git a/ovn/ovn-sb.ovsschema b/ovn/ovn-sb.ovsschema
> > index a9a91e5..ead733b 100644
> > --- a/ovn/ovn-sb.ovsschema
> > +++ b/ovn/ovn-sb.ovsschema
> > @@ -1,7 +1,7 @@
> > {
> > "name": "OVN_Southbound",
> > -"version"
Apart from some minor changes highlighted inline, looks good to me.
Acked-by: Sorin Vinturis
-Original Message-
From: dev [mailto:dev-boun...@openvswitch.org] On Behalf Of Alin Serdean
Sent: Saturday, 12 March, 2016 23:25
To: dev@openvswitch.org
Subject: [ovs-dev] [PATCH v2 1/3] windows:
The previous definitions of these variables using designated
initializers caused a variety of issues when attempting to
compile with MSVC, particularly if including these headers from C++
code. By defining them like this, we can appease MSVC and keep the
definitions the same on all platforms.
Sugg
Localnet port is now able to connect vif ports on different HVs.
Change the test case accordingly.
Signed-off-by: Han Zhou
---
Notes:
v1->v2:
Update according to Russell's comment: add test for connectivity
between 2 lswitches on same physical network
tests/ovn.at | 34
Thanks Ben! I will update the name as per your comment.
can you please answer 2nd part of the question? I have to write
encode,decode,parse and format functions for this new action. I have
updated decode and encode as per previously defined actions but find it
difficult for "parse" and format func
Hi Ben
Thank you for looking into this.
We are working on a generic tunneling offloading APIs and more updates will
follow.
Regards
_Sugesh
> -Original Message-
> From: Ben Pfaff [mailto:b...@ovn.org]
> Sent: Friday, March 18, 2016 2:53 AM
> To: Chandran, Sugesh
> Cc: dev@openvswitch.o
Recirculation support for the OVS extension.
Tested using PING and iperf with Driver Verifier enabled.
Signed-off-by: Sorin Vinturis
Co-authored-by: Alin Gabriel Serdean
Reported-by: Sorin Vinturis
Reported-at: https://github.com/openvswitch/ovs-issues/issues/104
---
v2: Initialize flow key be
Original Message
> From: Ben Warren
>
> Signed-off-by: Ben Warren
> ---
> FAQ.md | 2 +-
> build-aux/extract-ofp-errors | 4 +-
> include/openvswitch/automake.mk | 1 +
> include/openvswitch/ofp-errors.h | 807 +
On Wed, Mar 02, 2016 at 03:56:21PM +0100, Quentin Monnet wrote:
> Document the `--color` option for ovs-ofctl in its manpage. The option
> documentation has been added into a new file (included into the
> manpage) so that it can easily be added to other manpages as well when
> support for colorized
Recirculation feature probe messages sent from userspace are
suppressed by the OVS extension.
This patch changes the current behaviour to allow feature probe
for recirculation.
Signed-off-by: Sorin Vinturis
Acked-by: Sairam Venugopal
---
datapath-windows/ovsext/Flow.c | 9 -
1 file cha
Hi Han Zhou,
Just curious and not related to the bitwise_rscan().
Do you get a chance to know what this kernel symbol is?
>
> + 11.34% ovn-controller [kernel.kallsyms] [k] 0x8104f45a
>
you can try "cat /proc/kallsyms" and see where it locates.
> + 8.15% ovn-controller libc-2.1
On Sun, Mar 20, 2016 at 01:06:57PM +0800, Peng He wrote:
> Hi, Ben,
> You're correct. I think the principle is to cache as few bits as
> possible. In fact to
> differentiate two prefixes, you just need 1 more bits in positions [16,24].
That's close. The principle is for megaflows to match
On Sun, Mar 20, 2016 at 9:41 AM, William Tu wrote:
>
> Hi Han Zhou,
>
> Just curious and not related to the bitwise_rscan().
> Do you get a chance to know what this kernel symbol is?
Here is the report with kernel symbols resolved.
--- before optimization ---
+ 36.27% ovn-controller ovn-contr
Hello Panu,
Thanks for the comments. I have follow up question on the auto discovery of
DPDK install location.
As the DPDK install location can vary with distros, do OVS configure script
has to search for DPDK libraries in /usr/local/share as a starting point and
proceed to /usr/share locat
30+ Day Old Patches
===
ID State Date Name
-- -
584769 New2016-02-18 15:07:44 [ovs-dev,RFC] OVN: Quick implementation of
usage of conditional monitoring
585154 New2016-02-19 11:25:12 [ovs-dev,V5,2/2] netdev-d
On Fri, Mar 18, 2016 at 09:45:08AM -0700, Jarno Rajahalme wrote:
> OpenFlow 1.5 spec says: “The group type need not be specified for the delete
> request.”
Thanks Jarno. What I am wondering at this point is what
omitting the group type looks like on the wire.
_
Hi, Ben,
>We're always excited to improve the performance of OVS, so I hope you
>will pass along your results.
We did some evaluations on DPDK-based OVS. We use ClassBench[1] to
generate
1K and 10K rules, and also generate synthetic traffic for these rules. We
choose to generate
low loc
Hi Ryan,
I plan to test this patch in our scalability test environment, but it conflicts
with bce7cf on master, lfow_run()’s definition had been changed a lot.
Would it be possible to get a rebase of this patch on current master?
BR
Huang Lei
___
dev
25 matches
Mail list logo