Hi ALL:
We are running openstack J with neutron OVS2.4 using intel 10G
adapters(82599es) with vxlan tunnels.
We find this patch below to resolve problem of vxlan throughput.
http://openvswitch.org/pipermail/dev/2015-August/059335.html
We modify vport-vxlan.c as the patch,and mak
Please also see my comments for the next patch.
On Sat, Jan 16, 2016 at 12:16 AM, Liran Schour wrote:
> Indexing is different between ovsdb_row to ovsdb_monitor_row. To be
> able to evaluate conditions on ovsdb_monitor_row we need this mapping.
>
> signed-off-by: Liran Schour
> ---
> ovsdb/mon
Thank you, Aaron.
On 20.01.2016 19:41, Aaron Conole wrote:
> Ilya Maximets writes:
>> On 20.01.2016 18:37, Traynor, Kevin wrote:
>>>
>>> Hi Ilya, I'll review this when I'm finished reviewing Aaron's cmdline
>>> patchset.
>>
>> Thank you.
>>
>> But it is not only about this patch. Here is at leas
On Wed, Jan 20, 2016 at 3:22 PM, Ben Pfaff wrote:
>
> On Wed, Jan 20, 2016 at 11:31:08AM -0500, Russell Bryant wrote:
> > Multiple logical ports on the same chassis that were connected to the
> > same physical network via localnet ports were not able to send packets
> > to each other. This was be
On Wed, Jan 20, 2016 at 09:47:32PM -0800, Andy Zhou wrote:
> On Wed, Jan 20, 2016 at 4:33 PM, Ben Pfaff wrote:
>
> > Until now, ofputil_decode_switch_features() has put the ports from the
> > switch features message into a separate ofpbuf supplied as an argument.
> > The natural desire for a call
On Wed, Jan 20, 2016 at 4:33 PM, Ben Pfaff wrote:
> Until now, ofputil_decode_switch_features() has put the ports from the
> switch features message into a separate ofpbuf supplied as an argument.
> The natural desire for a caller is to just reuse an ofpbuf that it already
> has, and that's what
On 20 January 2016 at 21:17, Pravin B Shelar wrote:
> tun-dst could be NULL in case of incorrect action list
> where set tunnel action is missing but packet is sent
> to tunnel vport.
>
> Signed-off-by: Pravin B Shelar
Thanks for the fix.
Acked-by: Joe Stringer
tun-dst could be NULL in case of incorrect action list
where set tunnel action is missing but packet is sent
to tunnel vport.
Signed-off-by: Pravin B Shelar
---
datapath/linux/compat/gso.h | 5 -
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/datapath/linux/compat/gso.h b/datap
On Thu, Jan 21, 2016 at 11:44:34AM +0800, Shi Xin Ruan wrote:
> I search the VTEP information, all say that VTEP only can bridge a single
> virtual network to vlan network.
> Our use case want thst OVN can support static routes.
>
> And I also notice that vtep schema has tables like "Logical_Route
The purpose of this patch is to refactor the software offloads found in
the VXLAN and GRE code and also to refactor how the maximmum segment
size for a given NBL is obtained.
This patch introduces two functions OvsApplySWChecksumOnNB and OVSGetTcpMSS.
OVSGetTcpMSS - will return the mss found in a
I did not look at the details yet, but basically the reference would be
released if gotten initially and not taken for the flow install.
Jarno
> On Jan 20, 2016, at 4:59 PM, Ben Pfaff wrote:
>
> I think I like the idea of only needing a recirc_state in that code, but
> I don't think I fully
On 20 January 2016 at 16:13, Ansis Atteka wrote:
>
>
> On 20 January 2016 at 15:36, Ben Pfaff wrote:
>
>> On Wed, Jan 20, 2016 at 03:34:49PM -0800, Ben Pfaff wrote:
>> > On Wed, Jan 20, 2016 at 02:59:03PM -0800, Ansis Atteka wrote:
>> > > CentOS, RHEL and Fedora distributions ship with their own
Not a full review, but see some comments below,
And sorry for not snipping, but I felt like not cutting any context when
commenting an RFC patch.
Jarno
> On Jan 18, 2016, at 11:27 PM, Ben Pfaff wrote:
>
> One purpose of OpenFlow packet-in messages is to allow a controller to
> interpose on
Hi,
Thank you for your reply. I tried that before, but there is no ovs-dpdk in
the datapath folder. So I cannot execute the command below:
./datapath/dpdk/ovs-dpdk -c 0x0F -n 4 --proc-type primary -- --stats_core=0
--stats_int=5 -p 0x03
./vswitchd/ovs-vswitchd -c 0x100 --proc-type=secondary --
Thanks! I applied this to master.
On Thu, Jan 21, 2016 at 01:24:12AM +, Alin Serdean wrote:
> Acked-by: Alin Gabriel Serdean .
>
> > -Mesaj original-
> > De la: Ben Pfaff [mailto:b...@ovn.org]
> > Trimis: Thursday, January 21, 2016 3:19 AM
> > Către: dev@openvswitch.org
> > Cc: Ben P
Acked-by: Alin Gabriel Serdean .
> -Mesaj original-
> De la: Ben Pfaff [mailto:b...@ovn.org]
> Trimis: Thursday, January 21, 2016 3:19 AM
> Către: dev@openvswitch.org
> Cc: Ben Pfaff ; Alin Serdean
>
> Subiect: [PATCH] appveyor.yml: Install python "six" module.
>
> It is now required for
Anytime!
Alin.
> -Mesaj original-
> De la: Ben Pfaff [mailto:b...@ovn.org]
> Trimis: Thursday, January 21, 2016 3:22 AM
> Către: Alin Serdean
> Cc: dev@openvswitch.org
> Subiect: Re: [ovs-dev] python problems in Windows build--need help
>
> It does, thanks a lot for the help.
>
> I pos
It does, thanks a lot for the help.
I posted a fix:
http://openvswitch.org/pipermail/dev/2016-January/064803.html
On Thu, Jan 21, 2016 at 12:48:05AM +, Alin Serdean wrote:
> PS C:\Users\alin.cloudbase> pip install six --disable-pip-version-check
> Collecting six
> Using cached six-1
It is now required for the build.
CC: Alin Serdean
Signed-off-by: Ben Pfaff
---
appveyor.yml | 2 ++
1 file changed, 2 insertions(+)
diff --git a/appveyor.yml b/appveyor.yml
index 9efddeb..de0b4a5 100644
--- a/appveyor.yml
+++ b/appveyor.yml
@@ -33,6 +33,8 @@ init:
cd C:\openvswitch
+
Thanks, I folded it in.
On Wed, Jan 20, 2016 at 04:41:24PM -0800, Jarno Rajahalme wrote:
> I was a bit too fast, I had to add this incremental to make this patch
> compile:
>
> diff --git a/ofproto/ofproto-dpif-xlate.c b/ofproto/ofproto-dpif-xlate.c
> index 184eb46..6549191 100644
> --- a/ofprot
Thanks for the reviews. I've pushed the patches up to this point.
On Wed, Jan 20, 2016 at 04:39:10PM -0800, Jarno Rajahalme wrote:
> Acked-by: Jarno Rajahalme
>
> > On Jan 18, 2016, at 11:27 PM, Ben Pfaff wrote:
> >
> > The existing functions either assumed that we were working with NXM
> > (
I think I like the idea of only needing a recirc_state in that code, but
I don't think I fully understand yet. I guess it's pretty easy to make
xlate_in_init() take the reference and pass back the recirc_state
instead of the recirc_id_node, but where would the eventual release of
the reference hap
On Wed, Jan 20, 2016 at 03:49:26PM -0800, Jarno Rajahalme wrote:
> Nice simplification! With one bug fix below:
>
> Acked-by: Jarno Rajahalme
Thanks for the review.
> > On Jan 18, 2016, at 11:27 PM, Ben Pfaff wrote:
> >
> > In my opinion, this makes better sense for the stack, because it's no
PS C:\Users\alin.cloudbase> pip install six --disable-pip-version-check
Collecting six
Using cached six-1.10.0-py2.py3-none-any.whl
Installing collected packages: six
Successfully installed six-1.10.0
PS C:\Users\alin.cloudbase>
Hope this works.
Alin.
> -Mesaj original-
> De la: Ben Pf
I was a bit too fast, I had to add this incremental to make this patch compile:
diff --git a/ofproto/ofproto-dpif-xlate.c b/ofproto/ofproto-dpif-xlate.c
index 184eb46..6549191 100644
--- a/ofproto/ofproto-dpif-xlate.c
+++ b/ofproto/ofproto-dpif-xlate.c
@@ -5143,7 +5143,7 @@ xlate_actions(struct xl
Acked-by: Jarno Rajahalme
> On Jan 18, 2016, at 11:27 PM, Ben Pfaff wrote:
>
> The existing functions either assumed that we were working with NXM
> (instead of OXM), or they added or parsed a header before OXM headers.
> These aren't needed when OXM is encapsulated inside a property, as will b
Maybe ofproto-dpif-upcall.c could be refactored to not need the recirc_id_node
pointer. It is only used for getting the recirc ID and managing recirc_id_node
reference counting. If the reference counting could be moved upstream, the
pointer could be removed. But the idea was to take a reference
On Wed, Jan 20, 2016 at 03:32:58PM -0800, William Tu wrote:
> Testcase 909: ofproto-dpif - patch ports
> The ofpbuf->base is set to OFPBUF_STUB at
> ofputil_decode_switch_features()->ofpbuf_use_const()
> so ofpbuf_delete does not free it. Call trace is below:
> vconn_stream_recv(vconn-strea
OK.
The warning appears to kill the appveyor build. Do you know the
appropriate way to ignore it?
On Thu, Jan 21, 2016 at 12:18:41AM +, Alin Serdean wrote:
> It's ok.
>
> Until it is fixed upstream I think we should force 7.1.2 and just ignore the
> warning.
>
> Alin.
>
> > -Mesaj or
Until now, ofputil_decode_switch_features() has put the ports from the
switch features message into a separate ofpbuf supplied as an argument.
The natural desire for a caller is to just reuse an ofpbuf that it already
has, and that's what one of the callers did. This however has the
nonobvious eff
On 20 January 2016 at 15:52, pravin shelar wrote:
> On Wed, Jan 20, 2016 at 3:26 PM, Joe Stringer wrote:
>> The entire OVS_GSO_CB was not preserved when handling IP fragments,
>> leading to the following NULL pointer dereference in ovs_stt_xmit(). Fix
>> this in the fragmentation handling code by
It's ok.
Until it is fixed upstream I think we should force 7.1.2 and just ignore the
warning.
Alin.
> -Mesaj original-
> De la: Ben Pfaff [mailto:b...@ovn.org]
> Trimis: Thursday, January 21, 2016 2:16 AM
> Către: Alin Serdean
> Cc: dev@openvswitch.org
> Subiect: Re: [ovs-dev] python
Acked-by: Jarno Rajahalme
> On Jan 18, 2016, at 11:27 PM, Ben Pfaff wrote:
>
> An upcoming commit will make it possible to essentially serialize the
> recirculation state into an OpenFlow message. For that purpose, we can't
> sensibly pass a "struct ofproto *", but a randomly generated UUID wo
[adding the list back, hope that's OK]
Is it better to ignore this problem or the first one where pip requests
an upgrade? Either way, can you suggest a patch to avoid the problem?
Thanks a lot,
Ben.
On Thu, Jan 21, 2016 at 12:05:45AM +, Alin Serdean wrote:
> Hope this helps:
>
> It's a b
Thank you!
William
On Wed, Jan 20, 2016 at 3:54 PM, Ben Pfaff wrote:
> On Tue, Jan 12, 2016 at 02:24:53PM -0800, William Tu wrote:
> > This patch forces appending "parent_process_name(PID)" when invoking
> > ovs-vsctl, in order to assist debugging. The patch is for Linux only.
> > For example:
>
On 20 January 2016 at 15:36, Ben Pfaff wrote:
> On Wed, Jan 20, 2016 at 03:34:49PM -0800, Ben Pfaff wrote:
> > On Wed, Jan 20, 2016 at 02:59:03PM -0800, Ansis Atteka wrote:
> > > CentOS, RHEL and Fedora distributions ship with their own Open vSwitch
> > > SELinux policy that is too strict and pre
Acked-by: Jarno Rajahalme
> On Jan 18, 2016, at 11:27 PM, Ben Pfaff wrote:
>
> These functions actually hash or compare recirc_state structs, so they
> should be named that way; recirc_metadata is only a small subset of
> recirc_state.
>
> Signed-off-by: Ben Pfaff
> ---
> ofproto/ofproto-dpif
This is nice reduction of code duplication as well.
Acked-by: Jarno Rajahalme
> On Jan 18, 2016, at 11:27 PM, Ben Pfaff wrote:
>
> Instead reconstruct it using the action set, since we already have the
> logic to do that.
>
> This seems a little nicer because we don't have to "trust" the met
On Wed, Jan 20, 2016 at 06:29:58PM -0500, Russell Bryant wrote:
> On Wednesday, January 20, 2016, Ben Pfaff wrote:
>
> > On Wed, Jan 20, 2016 at 04:48:15PM -0500, Russell Bryant wrote:
> > > On this one, I used "Grudgingly-Acked-by: ". I'll submit a patch to
> > > CONTRIBUTING.md to document thi
On Wed, Jan 20, 2016 at 03:52:08PM -0800, William Tu wrote:
> Signed-off-by: William Tu
Applied to master, thank you!
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
Hi Ben,
Yes, I've tested it. Let me fix the copyright date and resubmit.
Regards,
William
On Wed, Jan 20, 2016 at 3:59 PM, Ben Pfaff wrote:
> On Wed, Jan 20, 2016 at 10:00:40AM -0800, William Tu wrote:
> > Create a new ovn/utilities/bugtool directory, add ovn.xml to bugtool
> > plugins, and add
Acked-by: Jarno Rajahalme
> On Jan 18, 2016, at 11:27 PM, Ben Pfaff wrote:
>
> During translation it makes some sense to concatenate these in a single
> array, but in my opinion it's conceptually better to separate them for
> the recirc_state; they are not naturally the same thing.
>
> Signed-
On Wed, Jan 20, 2016 at 10:00:40AM -0800, William Tu wrote:
> Create a new ovn/utilities/bugtool directory, add ovn.xml to bugtool
> plugins, and add ovn-nbctl show, ovn-sbctl show, and ovn-sbctl lflow-list.
>
> Signed-off-by: William Tu
Did you test it? If so then all I see here is a wrong cop
On Tue, Jan 12, 2016 at 02:24:53PM -0800, William Tu wrote:
> This patch forces appending "parent_process_name(PID)" when invoking
> ovs-vsctl, in order to assist debugging. The patch is for Linux only.
> For example:
> User adds br0 by "ovs-vsctl add-br0", the log shows:
> "ovs-vsctl (invo
I see that the next patch fixed this bug,
Jarno
> On Jan 20, 2016, at 3:49 PM, Jarno Rajahalme wrote:
>
> Nice simplification! With one bug fix below:
>
> Acked-by: Jarno Rajahalme mailto:ja...@ovn.org>>
>
>> On Jan 18, 2016, at 11:27 PM, Ben Pfaff wrote:
>>
>> In my opinion, this makes b
Signed-off-by: William Tu
---
tests/ofproto-dpif.at | 4 ++--
tests/test-sflow.c| 4 ++--
2 files changed, 4 insertions(+), 4 deletions(-)
diff --git a/tests/ofproto-dpif.at b/tests/ofproto-dpif.at
index 4c2a995..a5e07da 100644
--- a/tests/ofproto-dpif.at
+++ b/tests/ofproto-dpif.at
@@ -5104
On Wed, Jan 20, 2016 at 3:26 PM, Joe Stringer wrote:
> The entire OVS_GSO_CB was not preserved when handling IP fragments,
> leading to the following NULL pointer dereference in ovs_stt_xmit(). Fix
> this in the fragmentation handling code by preserving the whole CB.
>
> BUG: unable to handle kern
Nice simplification! With one bug fix below:
Acked-by: Jarno Rajahalme
> On Jan 18, 2016, at 11:27 PM, Ben Pfaff wrote:
>
> In my opinion, this makes better sense for the stack, because it's not
> a packet or a collection of bytes, it's an array of struct mf_subvalue.
> (I left it as an ofpbuf
On Wed, Jan 20, 2016 at 03:34:49PM -0800, Ben Pfaff wrote:
> On Wed, Jan 20, 2016 at 02:59:03PM -0800, Ansis Atteka wrote:
> > CentOS, RHEL and Fedora distributions ship with their own Open vSwitch
> > SELinux policy that is too strict and prevents Open vSwitch to work
> > normally out of the box.
On Wed, Jan 20, 2016 at 02:59:03PM -0800, Ansis Atteka wrote:
> CentOS, RHEL and Fedora distributions ship with their own Open vSwitch
> SELinux policy that is too strict and prevents Open vSwitch to work
> normally out of the box.
>
> As a solution, this patch introduces a new package which will
Testcase 909: ofproto-dpif - patch ports
The ofpbuf->base is set to OFPBUF_STUB at
ofputil_decode_switch_features()->ofpbuf_use_const()
so ofpbuf_delete does not free it. Call trace is below:
vconn_stream_recv(vconn-stream.c:155)
do_recv(vconn.c:640)
vconn_recv(vconn.c:598)
vcon
On Wednesday, January 20, 2016, Ben Pfaff wrote:
> On Wed, Jan 20, 2016 at 04:48:15PM -0500, Russell Bryant wrote:
> > On this one, I used "Grudgingly-Acked-by: ". I'll submit a patch to
> > CONTRIBUTING.md to document this new header.
>
> I can't figure out whether this is a joke.
>
It is a jo
On 20 January 2016 at 14:12, pravin shelar wrote:
> On Wed, Jan 20, 2016 at 12:41 PM, Joe Stringer wrote:
> It is safer to do the same while de-fragmenting packets (handle_fragments()).
Thanks, I rolled in your feedback and sent a v2.
___
dev mailing l
The entire OVS_GSO_CB was not preserved when handling IP fragments,
leading to the following NULL pointer dereference in ovs_stt_xmit(). Fix
this in the fragmentation handling code by preserving the whole CB.
BUG: unable to handle kernel NULL pointer dereference at 001c
IP: [] ovs_stt_
Hi Ilya,
Thanks for the patch. I like the general approach and the implementation.
Couple of things:
* This change should definitely be mentioned in NEWS, since it breaks
compatibility.
* I think INSTALL.DPDK.md should be updated too, since it contains
references to the old options.
* One mo
On Wed, Jan 20, 2016 at 05:04:24PM -0500, Russell Bryant wrote:
> On 01/20/2016 04:25 PM, Ben Pfaff wrote:
> > On Tue, Jan 12, 2016 at 02:45:47PM -0500, Russell Bryant wrote:
> >> In Python 2, dict.items(), dict.keys(), and dict.values() returned a
> >> list. dict.iteritems(), dict.iterkeys(), and
Hi Ilya,
sorry for the delay. I've reviewed this patch and I'll start reviewing
the other two right away.
Thanks,
Daniele
On 20/01/2016 08:41, "Aaron Conole" wrote:
>Ilya Maximets writes:
>> On 20.01.2016 18:37, Traynor, Kevin wrote:
>>>
>>> Hi Ilya, I'll review this when I'm finished revi
On Wed, Jan 20, 2016 at 11:31:08AM -0500, Russell Bryant wrote:
> Multiple logical ports on the same chassis that were connected to the
> same physical network via localnet ports were not able to send packets
> to each other. This was because ovn-controller created a single patch
> port between br
On Wed, Jan 20, 2016 at 11:31:07AM -0500, Russell Bryant wrote:
> Before this patch, physical.c build up the set of local datapaths for
> its own use. I'd like to use it in another module in a later patch, so
> pull it out of physical. It's now populated by the bindings module,
> since that seems
On Wed, Jan 20, 2016 at 11:31:06AM -0500, Russell Bryant wrote:
> There were a couple of checks that checked for a parent port as the
> field being non-NULL. We should treat an empty string the same as NULL
> for this field.
>
> Signed-off-by: Russell Bryant
Acked-by: Ben Pfaff
___
On Mon, Jan 18, 2016 at 10:45:51AM -0500, Russell Bryant wrote:
> Multiple logical ports on the same chassis that were connected to the
> same physical network via localnet ports were not able to send packets
> to each other. This was because ovn-controller created a single patch
> port between br
CentOS, RHEL and Fedora distributions ship with their own Open vSwitch
SELinux policy that is too strict and prevents Open vSwitch to work
normally out of the box.
As a solution, this patch introduces a new package which will "loosen"
up "openvswitch_t" SELinux domain so that Open vSwitch could op
On Mon, Jan 18, 2016 at 10:45:50AM -0500, Russell Bryant wrote:
> Before this patch, physical.c build up the set of local datapaths for
> its own use. I'd like to use it in another module in a later patch, so
> pull it out of physical. It's now populated by the bindings module,
> since that seems
On Wed, Jan 20, 2016 at 05:33:16PM -0500, Russell Bryant wrote:
> While debugging a broken OVN environment yesterday, the problem turned
> out to be invalid entries in the logical port addresses column. In
> particular, the following command had been used:
>
> $ ovn-nbctl lport-set-addresses lp
On Wed, Jan 20, 2016 at 04:48:15PM -0500, Russell Bryant wrote:
> On this one, I used "Grudgingly-Acked-by: ". I'll submit a patch to
> CONTRIBUTING.md to document this new header.
I can't figure out whether this is a joke.
___
dev mailing list
dev@open
We recently added a build requirement for the Python "six" library.
This isn't installed on appveyor by default, so it breaks. No problem,
right--just add a command to install it. So I added (see
https://ci.appveyor.com/project/blp/ovs-reviews/build/1.0.2):
python -m pip install six
to appv
This patch updates the formatting for the Logical_Port addresses column
in the show command output. Previously, output would look like:
addresses: 00:00:00:00:00:01 192.168.1.1 00:00:00:00:00:01 192.168.1.2
Now it looks like:
addresses: ["00:00:00:00:00:01 192.168.1.1", "00:00:00:00:00:01 1
While debugging a broken OVN environment yesterday, the problem turned
out to be invalid entries in the logical port addresses column. In
particular, the following command had been used:
$ ovn-nbctl lport-set-addresses lp0 MAC IP
instead of:
$ ovn-nbctl lport-set-addresses lp0 "MAC IP"
Thi
On 01/20/2016 05:00 PM, Ben Pfaff wrote:
> On Thu, Jan 14, 2016 at 11:03:14AM -0500, Russell Bryant wrote:
>> While debugging a broken OVN environment yesterday, the problem turned
>> out to be invalid entries in the logical port addresses column. In
>> particular, the following command had been u
On Wed, Jan 20, 2016 at 12:41 PM, Joe Stringer wrote:
> The entire OVS_GSO_CB was not preserved when performing fragmentation,
> leading to the following NULL pointer dereference in ovs_stt_xmit(). Fix
> this in the fragmentation handling code by preserving the whole CB.
>
> BUG: unable to handle
On 01/20/2016 05:04 PM, Ben Pfaff wrote:
> On Thu, Jan 14, 2016 at 03:10:20PM -0500, Russell Bryant wrote:
>> All of the ovn-controller-vtep tests were failing on my laptop due to an
>> unexpected message in the ovs-vswitchd log related to my VPN. This
>> setting resolves it and makes all tests pa
On Thu, Jan 14, 2016 at 03:10:20PM -0500, Russell Bryant wrote:
> All of the ovn-controller-vtep tests were failing on my laptop due to an
> unexpected message in the ovs-vswitchd log related to my VPN. This
> setting resolves it and makes all tests pass.
>
> Fixes: 0c1e8a7d637e ("ovn-controller-
On 01/20/2016 04:25 PM, Ben Pfaff wrote:
> On Tue, Jan 12, 2016 at 02:45:47PM -0500, Russell Bryant wrote:
>> In Python 2, dict.items(), dict.keys(), and dict.values() returned a
>> list. dict.iteritems(), dict.iterkeys(), and dict.itervalues() returned
>> an iterator.
>>
>> As of Python 3, dict.i
On Thu, Jan 14, 2016 at 11:03:15AM -0500, Russell Bryant wrote:
> This patch updates the formatting for the Logical_Port addresses column
> in the show command output. Previously, output would look like:
>
> addresses: 00:00:00:00:00:01 192.168.1.1 00:00:00:00:00:01 192.168.1.2
>
> Now it look
On Thu, Jan 14, 2016 at 11:03:14AM -0500, Russell Bryant wrote:
> While debugging a broken OVN environment yesterday, the problem turned
> out to be invalid entries in the logical port addresses column. In
> particular, the following command had been used:
>
> $ ovn-nbctl lport-set-addresses lp
On Wed, Jan 13, 2016 at 03:20:00PM +0800, Shi Xin Ruan wrote:
> OVN is a SDN solution, in some case, customer need to bridge or routing
> between a SDN network and a non-SDN network.
> As I know, OVN has support east-west distribute router, and OVS has support
> SNAT functionality.
>
> How to conn
On 01/20/2016 04:27 PM, Ben Pfaff wrote:
> On Tue, Jan 12, 2016 at 02:45:48PM -0500, Russell Bryant wrote:
>> Octal constants must be written as 0o077 instead of 0077 to be
>> compatible with both Python 2 and 3.
>>
>> Signed-off-by: Russell Bryant
>
> What a terrible syntax.
I do not disagree.
The schema constraints that OVS supports certainly aren't perfect.
There are plenty of situations where one can come up with better kinds
of constraints. Maybe you have come up with one. If you're going to
propose adding a new kind of constraint, then I think it would be
worthwhile to think throu
On Tue, Jan 12, 2016 at 02:45:48PM -0500, Russell Bryant wrote:
> Octal constants must be written as 0o077 instead of 0077 to be
> compatible with both Python 2 and 3.
>
> Signed-off-by: Russell Bryant
What a terrible syntax.
Acked-by: Ben Pfaff
___
On Tue, Jan 12, 2016 at 02:45:47PM -0500, Russell Bryant wrote:
> In Python 2, dict.items(), dict.keys(), and dict.values() returned a
> list. dict.iteritems(), dict.iterkeys(), and dict.itervalues() returned
> an iterator.
>
> As of Python 3, dict.iteritems(), dict.itervalues(), and dict.iterkey
On Tue, Jan 12, 2016 at 02:45:46PM -0500, Russell Bryant wrote:
> Python 2 had range() and xrange(). xrange() is more efficient, but
> behaves differently so range() was retained for compatibility. Python 3
> only has range() and it behaves like Python 2's xrange().
>
> Remove explicit use of xr
On Tue, Jan 12, 2016 at 02:45:45PM -0500, Russell Bryant wrote:
> Fix imports of xmlrpclib to be compatible with Python 3. Python 2 had
> xmlrpclib (client) and SimpleXMLRPCServer (server). In Python 3, these
> have been renamed to xmlrpc.client and xmlrpc.server.
>
> The solution implemented he
The entire OVS_GSO_CB was not preserved when performing fragmentation,
leading to the following NULL pointer dereference in ovs_stt_xmit(). Fix
this in the fragmentation handling code by preserving the whole CB.
BUG: unable to handle kernel NULL pointer dereference at 001c
IP: [] ovs_s
When the DPDK init function is called, it changes the executing thread's
CPU affinity to a single core specified in -c. This will result in the
userspace bridge configuration thread being rebound, even if that is not
the intent.
This change fixes that behavior by rebinding to the original thread
a
Currently, configuration of DPDK parameters is done via the command line
through a --dpdk **OPTIONS** -- command line argument. This has a number of
challenges, including:
* It must be the first option passed to ovs-vswitchd
* It breaks from the way most other things are configured in OVS
* It does
A previous change moved some commonly used arguments from commandline to
the database, and with it the ability to pass arbitrary arguments to
EAL. This change allows arbitrary eal arguments to be provided
via a new db entry 'other_config:dpdk-extra' which will tokenize the
string and add it to the
Existing DPDK integration is provided by use of command line options which
must be split out and passed to librte in a special manner. However, this
forces any configuration to be passed by way of a special DPDK flag, and
interferes with ovs+dpdk packaging solutions.
This commit delays dpdk initia
The user has control over the DPDK internal lcore coremask, but this
parameter can be autofilled with a bit more intelligence. If the user
does not fill this parameter in, we use the lowest set bit in the
current task CPU affinity. Otherwise, we will reassign the current
thread to the specified lco
"Traynor, Kevin" writes:
>> -Original Message-
>> From: Aaron Conole [mailto:acon...@redhat.com]
>> Sent: Monday, January 18, 2016 8:29 PM
>> To: dev@openvswitch.org
>> Cc: Flavio Leitner; Traynor, Kevin; Panu Matilainen; Zoltan Kiss
>> Subject: [PATCH v5 2/4] netdev-dpdk: Convert initiali
On Wed, Jan 20, 2016 at 11:14:57AM +0530, niti Rohilla wrote:
> It means that tun_src and tun_dst parameters in ovs cannot be configured
> from ODL, as there is no corresponding OXM field defined for tun_src and
> tun_dst in Openflow specifications. I have looked into the code, for
> tun_src and tu
Thanks, all patches now applied up to this point.
On Tue, Jan 19, 2016 at 05:22:44PM -0800, Jarno Rajahalme wrote:
> Acked-by: Jarno Rajahalme
>
> > On Jan 18, 2016, at 11:27 PM, Ben Pfaff wrote:
> >
> > All of the callers of hash_words() and hash_words64() actually find it
> > easier to pass
On Tue, Jan 19, 2016 at 05:09:30PM -0800, Jarno Rajahalme wrote:
> With two questions for clarification below,
>
> Acked-by: Jarno Rajahalme
Thanks, applied.
> > On Jan 18, 2016, at 11:27 PM, Ben Pfaff wrote:
> > -if (cookie) {
> > -pin->cookie = *cookie;
> > -}
> >
Create a new ovn/utilities/bugtool directory, add ovn.xml to bugtool
plugins, and add ovn-nbctl show, ovn-sbctl show, and ovn-sbctl lflow-list.
Signed-off-by: William Tu
---
ovn/utilities/automake.mk | 2 ++
ovn/utilities/bugtool/automake.mk | 11 ++
v2:
- Start a new ovn/utilities/bugtool directory.
- Add new ovn scripts and plugin, ovn.xml.
- Apply changes to utilities/bugtool/automake.mk to remove "ovn/" so that after
installation,
ovn.xml will be copied to the same directory as ovs, e.g.,
"/usr/share/openvswitch/bugtool-plugins/netw
I think I understood it when I implemented it, but it's been a long time
and apparently I didn't write it down...
On Tue, Jan 19, 2016 at 04:32:35PM -0800, Jarno Rajahalme wrote:
> I never understood this special case anyway,
>
> Acked-by: Jarno Rajahalme
>
> > On Jan 18, 2016, at 11:27 PM, Ben
Thanks for the reviews. I applied the patches up to this point now.
On Tue, Jan 19, 2016 at 04:33:34PM -0800, Jarno Rajahalme wrote:
> Acked-by: Jarno Rajahalme
>
> > On Jan 18, 2016, at 11:27 PM, Ben Pfaff wrote:
> >
> > An upcoming commit will use this library from ofp-util instead of ofpro
-Original Message-
From: dev on behalf of Alin Serdean
Date: Tuesday, January 12, 2016 at 11:30 PM
To: "dev@openvswitch.org"
Subject: [ovs-dev] [PATCH] datapath-windows: fix endless loop on reboot
>Testing under 2012 gave some more inisight on an old bug.
>
>If a PNP event with the valu
Well, it unifies both. I changed it to "Decode and print".
On Tue, Jan 19, 2016 at 04:23:46PM -0800, Jarno Rajahalme wrote:
> Maybe the title should be “ofp-print: Print all async config messages the
> same way.” instead?
>
> Acked-by: Jarno Rajahalme
>
> > On Jan 18, 2016, at 11:27 PM, Ben P
> -Original Message-
> From: dev [mailto:dev-boun...@openvswitch.org] On Behalf Of Aaron Conole
> Sent: Tuesday, January 19, 2016 1:28 PM
> To: Qiu, Michael
> Cc: dev@openvswitch.org; Flavio Leitner; Zoltan Kiss
> Subject: Re: [ovs-dev] [PATCH v3 3/3] netdev-dpdk: Autofill lcore coremask if
Thanks, I improved that and applied this to master.
On Tue, Jan 19, 2016 at 03:57:38PM -0800, Jarno Rajahalme wrote:
> With a small nit below,
>
> Acked-by: Jarno Rajahalme
>
> > On Jan 18, 2016, at 11:27 PM, Ben Pfaff wrote:
> >
>
> (snip)
>
> > +struct ofp14_async_prop {
> > +uint64_t
1 - 100 of 127 matches
Mail list logo