>>On Mon, Dec 04, 2017 at 08:16:47PM +, Bhanuprakash Bodireddy wrote:
>>> Processors support prefetch instruction in anticipation of write but
>>> compilers(gcc) won't use them unless explicitly asked to do so even
>>> with '-march=native' specified.
>>>
>>> [Problem]
>>> Case A:
>>>
> Prefetch the cacheline having the cycle stats so that we can speed up
> the cycles_count_start() and cycles_count_intermediate().
Do you have any performance results?
>
> Signed-off-by: Bhanuprakash Bodireddy
> ---
> lib/dpif-netdev.c | 3 ++-
> 1 file changed, 2 insertions(+), 1
Move tunnel neigh snooping from do_xlate_actions() to
terminate_native_tunnel() in order to keep ARP neighbor cache clean.
Furthermore filter ARP reply and Neighbor Advertisement messages
addressing tunnel endpoint.
Signed-off-by: Zoltan Balogh
---
If ARP reply is received on LOCAL port of underlay bridge, then packets
pushed to tunnel port in overlay bridge will be dropped in the underlay.
This commit modifies this by receiving ARP replies on a different dummy
port. This fix is needed for an upcomming correction for ARP snooping.
By avoiding Tx recirculation and embracing tnl_push action within clone,
the tunnel metadata is not propagated. Unless clone action is handled in
the dpif_sflow_read_actions() function as well. This commit resolves
this issue.
In addition, some sflow data needs to be stored and restored in
Currenlty, OVS snoops any ARP or ND packets in any bridge and populates
the tunnel neighbor cache with the retreived data. For instance, when
ARP reply originated by a tenant is received on an overlay bridge, the
ARP message is snooped and tunnel neighbor cache is filled with tenant
data, however
> On Nov 21, 2017, at 5:00 PM, Yi-Hung Wei wrote:
>
> This patch adds support of flushing a conntrack entry specified by the
> conntrack 5-tuple, and provides the implementation in dpif-netlink.
> The implementation of dpif-netlink in the linux datapath utilizes the
>
added a unit test case for testing the condition when a tap interface
is added to br0 and then the tap interface is deleted from the
system and added back with the same name.
Signed-off-by: Ashish Varma
---
tests/automake.mk | 3 ++-
>> On Nov 21, 2017, at 5:00 PM, Yi-Hung Wei wrote:
>>
>> This patch adds support of flushing a conntrack entry specified by the
>> conntrack 5-tuple, and provides the implementation in dpif-netlink.
>> The implementation of dpif-netlink in the linux datapath utilizes the
>>
[...]
>int main()
>{
>int c;
>
>__builtin_prefetch(, 1, 1);
>c = 8;
>
>return c;
>}
>
>on my old Ivy Bridge i7-3770 CPU. It does not support even 'prefetchw':
>
> PREFETCHWT1 = false
> 3DNow! PREFETCH/PREFETCHW instructions =
Lucas asked me about backporting this one, as OpenStack would start
making use of it with an OVS 2.8 update if available.
The schema change seems pretty harmless. The catch is that this also
updated the schema version number from "5.8.1" to "5.8.2", while
branch-2.8 has "5.8.0". master includes
We have now repeated our earlier iperf3 tests for this patch series.
https://mail.openvswitch.org/pipermail/ovs-dev/2017-September/338247.html
We use an iperf3 server as representative for a typical IO-intensive kernel
application. The iperf3 server executes in a VM with 2 vCPUs where both
On Tue, Oct 31, 2017 at 3:55 PM, Ben Pfaff wrote:
> On Thu, Aug 17, 2017 at 02:14:13PM -0500, Terry Wilson wrote:
>> The JSON C extensions performs much better than the pure Python
>> version, so build it when producing RPMs.
>>
>> Signed-off-by: Terry Wilson
>
If you are not able to see this mail, click
http://r.newsletter.bonescamail.nl/7xa28ko66oatrf.html
[
]( http://r.newsletter.bonescamail.nl/track/click/vp48yaoodaoatrd )
TODAY'S NEW ARRIVAL
TOPQALITY JONA BLACK TIGER SHRIMPS BANGLADESH 10 X 1 KILO / 700 GRS NETTO
THAWED WEIGHT AVERAGE
On Dec 5, 2017 9:16 PM, "Russell Bryant" wrote:
On Mon, Dec 4, 2017 at 9:27 AM, wrote:
> From: Numan Siddique
>
> Pacemaker Resource agent periodically calls the OVN OCF's "monitor" action
> periodically to check the status. But the
On Tue, Dec 5, 2017 at 1:17 PM, Numan Siddique wrote:
>
>
> On Dec 5, 2017 9:16 PM, "Russell Bryant" wrote:
>
> On Mon, Dec 4, 2017 at 9:27 AM, wrote:
>> From: Numan Siddique
>>
>> Pacemaker Resource agent
I don't object to #2.
On Tue, Dec 05, 2017 at 01:44:25PM -0500, Russell Bryant wrote:
> Lucas asked me about backporting this one, as OpenStack would start
> making use of it with an OVS 2.8 update if available.
>
> The schema change seems pretty harmless. The catch is that this also
> updated
An extended netlink ack has been added for 4.14 - add compat layer
changes so that it compiles for all kernels up to and including
4.14.
Signed-off-by: Greg Rose
---
acinclude.m4| 3 +++
datapath/linux/compat/include/linux/netdevice.h |
Various fixes and compat layer changes required to enable building
OVS for the upstream Linux 4.14 kernel.
The constant changing of the netdev_master_upper_dev_link parameters
is a real headache. I couldn't think of any cleaner way to do it
than the approach I used but I welcome suggestions on
It is more portable.
Signed-off-by: Ben Pfaff
---
tests/automake.mk | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/tests/automake.mk b/tests/automake.mk
index c5cc95251207..ad5362ff1f5b 100644
--- a/tests/automake.mk
+++ b/tests/automake.mk
@@ -167,7 +167,7 @@
"test -d x || mkdir x" has a race when invoked in parallel: it is possible
for two processes to both see that 'x' does not exist and both try to
create it, and if that happens then one of them will fail. This avoids
the problem.
Signed-off-by: Ben Pfaff
---
tests/automake.mk | 2
Signed-off-by: Greg Rose
---
.travis.yml | 17 -
1 file changed, 8 insertions(+), 9 deletions(-)
diff --git a/.travis.yml b/.travis.yml
index f217840..8ca6d75 100644
--- a/.travis.yml
+++ b/.travis.yml
@@ -25,17 +25,16 @@ sudo: false
env:
-
Una útil herramienta jurídica para sus necesidades personales y comerciales vial
Fideicomisos en México
11 de Diciembre - Dr. Eduardo Rocha Núñez 9am-6pm
Usted puede obtener numerosos beneficios económicos con la constitución e
implementación de un fideicomiso, un instrumento legislativo y
Using SKB_GSO_UDP breaks the compilation on Linux 4.14. Check for
the HAVE_SKB_GSO_UDP compiler #define.
Signed-off-by: Greg Rose
---
datapath/datapath.c | 9 ++---
datapath/linux/compat/stt.c | 11 ++-
2 files changed, 16 insertions(+), 4 deletions(-)
Linux 4.14 requires this parameter to be constified.
Signed-off-by: Greg Rose
---
datapath/conntrack.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/datapath/conntrack.c b/datapath/conntrack.c
index d517a87..4b181ad 100644
--- a/datapath/conntrack.c
If the entire file is not going to be compiled due to compilation time
directives then don't bother pulling in the headers. This fixes a
compilation error on Linux 4.14.
Signed-off-by: Greg Rose
---
datapath/linux/compat/ip_gre.c| 2 +-
This feature is used to limit the bandwidth of flows, such as floating IP.
ovn-northd changes:
1. add bandwidth column in NB's QOS table.
2. add QOS_METER stages in Logical switch ingress/egress.
3. add set_meter() action in SB's LFlow table.
ovn-controller changes:
add meter_table for meter
The structure and function of the group table and meter table are similar,
refactoring code is used to extend for add the meter table.
The following function as lib: table init/destroy/clear,
install contents from desired, remove contents from existing,
Move the contents of desired to existing.
Guoshuai Li (2):
ovn-controller: Add extend_table instead of group_table to expand
meter.
ovn: OVN Support QoS meter
NEWS| 1 +
include/ovn/actions.h | 34 +++--
ovn/controller/lflow.c | 18 ++-
ovn/controller/lflow.h | 5 +-
On Mon, Dec 4, 2017 at 9:27 AM, wrote:
> From: Numan Siddique
>
> Pacemaker Resource agent periodically calls the OVN OCF's "monitor" action
> periodically to check the status. But the OVN OCF script doesn't add the
> action "monitor" for the role
>
>> Prefetch the cacheline having the cycle stats so that we can speed up
>> the cycles_count_start() and cycles_count_intermediate().
>
>Do you have any performance results?
I don’t have nos. for this patch alone. I was testing the overall throughput
along with other patches (that were *not*
On 05.12.2017 16:54, Bodireddy, Bhanuprakash wrote:
On Mon, Dec 04, 2017 at 08:16:47PM +, Bhanuprakash Bodireddy
>> wrote:
> Processors support prefetch instruction in anticipation of write but
> compilers(gcc) won't use them unless explicitly asked to do so even
> with
- New get_custom_stats interface function is added to netdev. It
allows particular netdev implementation to expose custom
counters in dictionary format (counter name/counter value).
- New statistics are retrieved using experimenter code and
are printed as a result to ofctl dump-ports.
- New
>I have retested your "Output patches batching" v6 in our standard PVP L3-
>VPN/VXLAN benchmark setup [1]. The configuration is a single PMD serving a
>physical 10G port and a VM running DPDK testpmd as IP reflector with 4
>equally loaded vhostuser ports. The tests are run with 64 byte packets.
>>>On Mon, Dec 04, 2017 at 08:16:47PM +, Bhanuprakash Bodireddy
>wrote:
Processors support prefetch instruction in anticipation of write but
compilers(gcc) won't use them unless explicitly asked to do so even
with '-march=native' specified.
[Problem]
Case A:
35 matches
Mail list logo