Hi Jesse
Thanks for responding.
I checked this earlier and it does not add up. I looked at the delta before and
after I ran the netperf test for 10s on a 10G link ( I get ~6G with my current
parameters). The ifconfig numbers are in the ballpark for 10s of transmission
at ~6G, but the ovs-dpctl
On Wed, Jun 5, 2013 at 3:24 AM, Maciej Gałkiewicz
wrote:
> The problem with version 1.9.0 is that it does not support kernel greater
> than 3.8. On the other way kernel 3.9.4 is already in sid.
No version of Open vSwitch currently supports Linux 3.9. As an out of
tree module, there is no general
On Fri, Jun 7, 2013 at 11:46 AM, K.R Kishore wrote:
> Hi
> I am trying to understand the statistics from ovs-dpctl and reconcile that
> with stats from other tools. I am finding it difficult to make sense of the
> numbers, perhaps someone can point me to some documentation about how this
> works?
OK, here's the second half:
On Wed, Jun 5, 2013 at 10:46 PM, Andy Zhou wrote:
> diff --git a/datapath/flow.c b/datapath/flow.c
> index 7f897bd..fc8fb86 100644
> --- a/datapath/flow.c
> +++ b/datapath/flow.c
> static struct kmem_cache *flow_cache;
> +static LIST_HEAD(mask_list);
We should add a
Acked-by: Ansis Atteka
On Fri, Jun 7, 2013 at 10:14 AM, Ben Pfaff wrote:
> Signed-off-by: Ben Pfaff
> ---
> lib/ofp-actions.h | 12 ++--
> 1 files changed, 6 insertions(+), 6 deletions(-)
>
> diff --git a/lib/ofp-actions.h b/lib/ofp-actions.h
> index 9a74bcc..4e7fb5c 100644
> --- a
On Wed, Jun 5, 2013 at 1:05 PM, Ben Pfaff wrote:
> Returning a static data buffer makes code more brittle and definitely
> not thread-safe, so this commit switches to using a caller-provided
> buffer instead.
>
> Signed-off-by: Ben Pfaff
>
> Acked-by: Ansis Atteka
__
I've uploaded a first cut of an updated version of Open vSwitch for
Debian for review:
http://mentors.debian.net/package/openvswitch
Please feel free to test and comment on it, either on the mailing list
here or in the comment box on that webpage.
__
I think it should go into Interface.
The field should be mutable, because its value is not known when
ovs-vsctl (or another OVSDB client) adds the Interface record;
instead, ovs-vswitchd fills it in later. Otherwise your plan sounds
right.
Other interface status columns, such as 'link_state' and
Want to make sure I get off on the right foot... should the read-only ifindex
field go
into "Port" or "Interface" ? It seems like ovsrec_interface is the one with
the pointer to netdev.
So I just need to:
1. change the schema to add ifindex as a new field (with mutable: false),
2. update the
Some tools do not like text files that lack a trailing new-line. In
particular, Debian's dpkg-source utility complains about a missing new-line
in the file generated by ovsdb-idlc:
dpkg-source: warning: file
openvswitch-1.9.2+git20130605/lib/vswitch-idl.ovsidl has no final
newline (ei
On Wed, Jun 5, 2013 at 1:05 PM, Ben Pfaff wrote:
> Returning a static data buffer makes code more brittle and definitely
> not thread-safe, so this commit switches to using a caller-provided
> buffer instead.
>
> Signed-off-by: Ben Pfaff
> ---
> lib/ofp-actions.c |5 -
> lib/ofp-print.c
On Wed, Jun 5, 2013 at 1:05 PM, Ben Pfaff wrote:
> Signed-off-by: Ben Pfaff
> ---
> lib/netdev-vport.c |7 ++-
> 1 files changed, 2 insertions(+), 5 deletions(-)
>
> diff --git a/lib/netdev-vport.c b/lib/netdev-vport.c
> index 3558f43..4bb41bd 100644
> --- a/lib/netdev-vport.c
> +++ b/li
Thanks merged.
Ethan
On Fri, Jun 7, 2013 at 1:58 PM, Ben Pfaff wrote:
> On Fri, Jun 07, 2013 at 01:55:32PM -0700, Ethan Jackson wrote:
>> In almost all cases, ovs-vswitchd runs with a single datapath, and
>> therefore a single governor. Therefore, it's usually clear from
>> context what governo
On Fri, Jun 07, 2013 at 01:55:32PM -0700, Ethan Jackson wrote:
> In almost all cases, ovs-vswitchd runs with a single datapath, and
> therefore a single governor. Therefore, it's usually clear from
> context what governor log messages are referring to making the name
> redundant. This patch remov
In almost all cases, ovs-vswitchd runs with a single datapath, and
therefore a single governor. Therefore, it's usually clear from
context what governor log messages are referring to making the name
redundant. This patch removes it.
Signed-off-by: Ethan Jackson
---
ofproto/ofproto-dpif-governo
On Fri, Jun 07, 2013 at 01:48:31PM -0700, Ethan Jackson wrote:
> > The incremental is just ofproto-dpif.c, did you intend to remove
> > 'name' from struct governor in ofproto-dpif-governor.h also?
>
> Ah that really should be a separate patch IMO. How about I merge this
> series and then send som
> The incremental is just ofproto-dpif.c, did you intend to remove
> 'name' from struct governor in ofproto-dpif-governor.h also?
Ah that really should be a separate patch IMO. How about I merge this
series and then send something out which changes the governor.
Ethan
___
Right, sorry for the mix up,
Jarno
On Jun 7, 2013, at 23:38 , ext Ben Pfaff wrote:
> Are we talking about the datapath or userspace? The datapath would
> use the Linux kernel roundup(), from linux/kernel.h. Userspace would
> use ROUND_UP(), from lib/util.h.
>
> On Fri, Jun 07, 2013 at 08:28
On Fri, Jun 07, 2013 at 01:37:48PM -0700, Ethan Jackson wrote:
> > Practically speaking, there is generally one backer. So I think that
> > places where previously the code logged a bridge name, and now logs a
> > backer name, we might as well drop the name entirely. This also goes
> > for the go
> Whoa. Declaring a static variable in a header file? Far out. But it
> isn't going to work.
Doh you're right. I'll just make it an extern. See the incremental below.
Looks alright otherwise?
---
ofproto/ofproto-provider.h |3 +--
ofproto/ofproto.c |2 ++
2 files changed, 3
> Practically speaking, there is generally one backer. So I think that
> places where previously the code logged a bridge name, and now logs a
> backer name, we might as well drop the name entirely. This also goes
> for the governor, which doesn't seem to need a name at all anymore.
Fine with me
Are we talking about the datapath or userspace? The datapath would
use the Linux kernel roundup(), from linux/kernel.h. Userspace would
use ROUND_UP(), from lib/util.h.
On Fri, Jun 07, 2013 at 08:28:20PM +, Rajahalme, Jarno (NSN - FI/Espoo)
wrote:
> It is in lib/util.h:
>
> lib/util.h:#def
On Fri, Jun 07, 2013 at 01:32:12PM -0700, Ethan Jackson wrote:
> > Whoa. Declaring a static variable in a header file? Far out. But it
> > isn't going to work.
>
> Doh you're right. I'll just make it an extern. See the incremental below.
> Looks alright otherwise?
Yes, thanks.
__
This patch is required for subsequent mega flow patchs.
---
include/linux/openvswitch.h |9 -
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/include/linux/openvswitch.h b/include/linux/openvswitch.h
index e890fd8..a119b14 100644
--- a/include/linux/openvswitch.h
+++ b/inc
It is in lib/util.h:
lib/util.h:#define DIV_ROUND_UP(X, Y) (((X) + ((Y) - 1)) / (Y))
lib/util.h:#define ROUND_UP(X, Y) (DIV_ROUND_UP(X, Y) * (Y))
Jarno
On Jun 7, 2013, at 17:39 , ext Andy Zhou wrote:
Thanks for the suggestion.
The kernel source I am looking at: 3.9.0-rc8, does not a accessib
Accept Invitation
http://www.facebook.com/r.php?re=d1875e3cc3a00a2a2e478494fd7aeea7&mid=HMTM2MDY5MDAyOmRldkBvcGVudnN3aXRjaC5vcmc6NzA&key=AQDnyVLH5mEVJn-SNPP9B_EgjJjtQZWEzuGoBbY9-e5HBL9zG_SPbq7zdmgYvamw5ruERteQM4xC8qtSLdqClVTXW14&invid=16077314184
Go to
Agreed, It's not particularly complicated. Intended to make the
future patches easier to review.
Ethan
On Thu, Jun 6, 2013 at 2:59 PM, Ben Pfaff wrote:
> On Wed, Jun 05, 2013 at 01:46:19PM -0700, Ethan Jackson wrote:
>> Subfacets being per-datapath entities, their statistics are really
>> only
On Fri, Jun 07, 2013 at 01:22:19PM -0700, Ethan Jackson wrote:
> Here's an incremental. Also I've changed the commit message to the following:
All good, thanks.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
Here's an incremental. Also I've changed the commit message to the following:
ofproto-dpif: More responsive average subfacet statistics.
Before this patch, statistics about subfacet average life span,
and count in the datapath, were calculated over the entire lifetime
of the ofproto. Furthermor
On Fri, Jun 07, 2013 at 02:03:40AM -0700, Justin Pettit wrote:
> Dynamically determines the flow fields that were relevant in
> processing flows based on the OpenFlow flow table and switch
> configuration. The immediate use for this functionality is to
> generate an action translation cache for si
Title: dev
dev-优*惠开具企业正.规税..*票，可验*证后付*款。dev联系：137 1381 9792(林 *先* 生)
Q Q：1289 6006烦请保留联**系方式，以备急需！dev___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
Hi
I am trying to understand the statistics from ovs-dpctl and reconcile that with
stats from other tools. I am finding it difficult to make sense of the numbers,
perhaps someone can point me to some documentation about how this works? There
are numerous links on the web and it is quite confusin
Signed-off-by: Ben Pfaff
---
lib/ofp-actions.h | 12 ++--
1 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/lib/ofp-actions.h b/lib/ofp-actions.h
index 9a74bcc..4e7fb5c 100644
--- a/lib/ofp-actions.h
+++ b/lib/ofp-actions.h
@@ -385,14 +385,14 @@ struct ofpact_resubmit {
Verleiht private Darlehen kurzfristiger und langer gehender von 1000 Euro an
200 Euro an allen ernsthaften Personen, die ein Darlehen wünschen.
Kontaktieren Sie es durch E - Mail: t...@financier.com
Octroie
On Fri, Jun 7, 2013 at 3:35 AM, Andy Zhou wrote:
> On Thu, Jun 6, 2013 at 7:01 PM, Jesse Gross wrote:
>> On Wed, Jun 5, 2013 at 10:46 PM, Andy Zhou wrote:
>> > diff --git a/datapath/datapath.c b/datapath/datapath.c
>> > index 42af315..98d78a8 100644
>> > --- a/datapath/datapath.c
>> > +++ b/data
It looks to me like it is available all the way back to 2.6.18, which
is the oldest kernel we care about.
On Fri, Jun 7, 2013 at 7:39 AM, Andy Zhou wrote:
> Thanks for the suggestion.
>
> The kernel source I am looking at: 3.9.0-rc8, does not a accessible
> definition of ROUND_UP, but roundup(x,
On Fri, Jun 07, 2013 at 02:03:38AM -0700, Justin Pettit wrote:
> From: Ethan Jackson
>
> This function will be useful in a future commit.
>
> Signed-off-by: Ethan Jackson
> Co-authored-by: Justin Pettit
> Signed-off-by: Justin Pettit
Looks good, thanks.
__
On Fri, Jun 07, 2013 at 02:03:39AM -0700, Justin Pettit wrote:
> From: Ethan Jackson
>
> A future commit will want to know what bits were significant during the
> classifier lookup.
>
> Signed-off-by: Ethan Jackson
> Co-authored-by: Justin Pettit
> Signed-off-by: Justin Pettit
Looks good, th
On Fri, Jun 07, 2013 at 02:03:37AM -0700, Justin Pettit wrote:
> An obvious copy/paste error.
>
> Signed-off-by: Justin Pettit
Looks good, thanks.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Fri, Jun 07, 2013 at 01:51:26AM -0700, Brian Shang wrote:
> I want to add a match field to ovs, but I'm not sure whether it should be
> modified by "set_field" action. From openflow spec 1.3, it says "set_field"
> action can modify any valid OXM header field, but I find some fields are
> set fal
On Fri, Jun 07, 2013 at 02:47:49PM +0300, Lori Jakab wrote:
> On 6/7/13 3:18 AM, Ben Pfaff wrote:
> >Have you considered the OpenFlow implications? OpenFlow itself is
> >somewhat Ethernet centric. Improving this has been in discussion, in
> >various ways, in the "extensibility" working group at O
Thanks for the suggestion.
The kernel source I am looking at: 3.9.0-rc8, does not a accessible
definition of ROUND_UP, but roundup(x,y) is, and fits the bill. I am
thinking of using it instead, unless it breaks on some older kernel version
that we care about.
>From kernel.h:
/* The `const' in
On Jun 7, 2013, at 13:35 , ext Andy Zhou wrote:
Is this the same as DIV_ROUND_UP?
I did not use it because DIV_ROUND_UP gives the roundup of u32, I needed the
roundup in bytes. However, I could rewrite this function using DIV_ROUND_UP.
There is ROUND_UP doing that already.
On 05/27/13 at 10:28am, Jesse Gross wrote:
> On Saturday, May 25, 2013, Eric Dumazet wrote:
>
> > On Sat, 2013-05-25 at 08:02 +0100, Thomas Graf wrote:
> >
> > > I ran TCP_CRR to verify the SYN/ACK use case and I did not
> > > observe a difference. If you have any specific test in mind
> > > I wil
On 6/7/13 3:18 AM, Ben Pfaff wrote:
On Thu, Jun 06, 2013 at 12:56:35AM +0300, Lori Jakab wrote:
The LISP tunneling support as of now is not yet ready for upstreaming,
for reasons outlined in this message:
http://openvswitch.org/pipermail/dev/2013-February/025459.html
One solution to the above
Thanks for the review.
On Thu, Jun 6, 2013 at 7:01 PM, Jesse Gross wrote:
> On Wed, Jun 5, 2013 at 10:46 PM, Andy Zhou wrote:
> > Add mega flow support in kernel datapath.
> >
> > Pravin has made significant contributions to this patch. Including
> > the mega flow id look up scheme, API clean u
On Jun 7, 2013, at 12:03 , ext Justin Pettit wrote:
...
> diff --git a/lib/flow.c b/lib/flow.c
> index 6476029..db4ce1d 100644
> --- a/lib/flow.c
> +++ b/lib/flow.c
> @@ -759,6 +759,32 @@ flow_hash_symmetric_l4(const struct flow *flow, uint32_t
> basis)
> return jhash_bytes(&fields, sizeof fi
Click to view Web Version.
Women
An obvious copy/paste error.
Signed-off-by: Justin Pettit
---
lib/meta-flow.c |5 ++---
1 files changed, 2 insertions(+), 3 deletions(-)
diff --git a/lib/meta-flow.c b/lib/meta-flow.c
index 54bc4c2..f9e044d 100644
--- a/lib/meta-flow.c
+++ b/lib/meta-flow.c
@@ -1356,9 +1356,8 @@ mf_set_valu
From: Ethan Jackson
A future commit will want to know what bits were significant during the
classifier lookup.
Signed-off-by: Ethan Jackson
Co-authored-by: Justin Pettit
Signed-off-by: Justin Pettit
---
lib/classifier.c| 17 +++--
lib/classifier.h|3 ++-
ofp
From: Ethan Jackson
This function will be useful in a future commit.
Signed-off-by: Ethan Jackson
Co-authored-by: Justin Pettit
Signed-off-by: Justin Pettit
---
lib/flow.c | 43 +--
lib/flow.h |3 +++
2 files changed, 32 insertions(+), 14 deletio
On Jun 6, 2013, at 2:02 PM, Ben Pfaff wrote:
> Every caller of lookup_xc() creates a new xc entry on failure. Can
> any of this be factored into a helper? If not, then all the callers
> of create_xc() use very similar 3 lines of code, so can that be
> integrated into create_xc()?
I think the i
Hi all,
I want to add a match field to ovs, but I'm not sure whether it should be
modified by "set_field" action. From openflow spec 1.3, it says "set_field"
action can modify any valid OXM header field, but I find some fields are
set false(eg. "ipv6_label"). I want to know how can we determaine it
53 matches
Mail list logo
