- Original Message -
> From: "Jesse Gross"
> To: "Romain Lenglet"
> Cc: "Neil Mckee" , dev@openvswitch.org
> Sent: Friday, October 18, 2013 6:23:23 PM
> Subject: Re: [ovs-dev] [PATCH] sFlow export: include standard tunnel
> structures (for GRE, VXLAN etc.)
>
> On Fri, Oct 18, 2013 at 5:
On Fri, Oct 18, 2013 at 5:58 PM, Romain Lenglet wrote:
> - Original Message -
>> From: "Jesse Gross"
>> To: "Romain Lenglet"
>> Cc: "Neil Mckee" , dev@openvswitch.org
>> Sent: Friday, October 18, 2013 5:50:05 PM
>> Subject: Re: [ovs-dev] [PATCH] sFlow export: include standard tunnel
>>
Hi Neil,
Sorry for the long delay to answer.
- Original Message -
> From: "Neil Mckee"
> To: "Jesse Gross"
> Cc: dev@openvswitch.org
> Sent: Wednesday, October 9, 2013 12:33:35 PM
> Subject: Re: [ovs-dev] [PATCH] sFlow export: include standard tunnel
> structures (for GRE, VXLAN etc.)
- Original Message -
> From: "Jesse Gross"
> To: "Romain Lenglet"
> Cc: "Neil Mckee" , dev@openvswitch.org
> Sent: Friday, October 18, 2013 5:50:05 PM
> Subject: Re: [ovs-dev] [PATCH] sFlow export: include standard tunnel
> structures (for GRE, VXLAN etc.)
>
> On Fri, Oct 18, 2013 at 5:
On Fri, Oct 18, 2013 at 5:43 PM, Romain Lenglet wrote:
> - Original Message -
>> From: "Romain Lenglet"
>> To: "Neil Mckee"
>> Cc: dev@openvswitch.org
>> Sent: Wednesday, October 9, 2013 10:30:17 AM
>> Subject: Re: [ovs-dev] [PATCH] sFlow export: include standard tunnel
>> structures (
On Fri, Oct 18, 2013 at 5:22 PM, Pravin B Shelar wrote:
> diff --git a/datapath/flow.c b/datapath/flow.c
> index faa4e15..73c4dbe 100644
> --- a/datapath/flow.c
> +++ b/datapath/flow.c
> +void ovs_flow_stats_get(struct sw_flow *flow, struct sw_flow_stats *res)
> +{
[...]
> + if (cpu
On Fri, Oct 18, 2013 at 4:54 PM, Andy Zhou wrote:
> Collect mega flow mask stats. ovs-dpctl show command can be used to
> display them.
>
> Signed-off-by: Andy Zhou
I noticed a few things that I was originally planning on just
correcting but it was a little more than I usually would like to do.
In current code, the odp_execute_actions() function does not check
the value of "userspace" function pointer before invoking it. This
patch adds a check for it.
Signed-off-by: Alex Wang
---
lib/odp-execute.c |4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/lib/odp-execu
Acked-by: Justin Pettit
--Justin
On Oct 18, 2013, at 4:30 PM, Ben Pfaff wrote:
> Commit 15aaf59932a3 (ofproto: Add global locking around flow table
> changes.) introduced doubled messages for expirations.
>
> Reported-by: John Hurley
> Signed-off-by: Ben Pfaff
> ---
> ofproto/ofproto.c |
With mega flow implementation ovs flow can be shared between
multiple CPUs which makes stats updates highly contended
operation. Following patch allocates separate stats for each
CPU to make stats update scalable.
Signed-off-by: Pravin B Shelar
---
v2:
disable bh for local stats read.
v3:
disable
Collect mega flow mask stats. ovs-dpctl show command can be used to
display them.
Signed-off-by: Andy Zhou
---
v1-v2:
* remove get_dp_megaflow_stats(), extend get_dp_stats() instead
* use __always_unused instead of GCC extensions
* fix comments
* fix the logic of
Signed-off-by: Andy Zhou
v1-v2:
* format fixes
* added contents to ovs-dpctl man page
* allow mask count to be zero.
v2->v3:
* Avoid divide by zero if n_pkts is zero.
---
NEWS |2 ++
lib/dpif-linux.c | 18 +-
Commit 15aaf59932a3 (ofproto: Add global locking around flow table
changes.) introduced doubled messages for expirations.
Reported-by: John Hurley
Signed-off-by: Ben Pfaff
---
ofproto/ofproto.c | 10 +-
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/ofproto/ofproto.c b/
Make union ofp_action more generic by allowing it to be used for
all actions. OFP_ACTION_ALIGN is now used where sizeof union
ofp_action was used before. On the same note, variable name
'n_actions' is changed to 'max_actions' as the number of actions
cannot be simply divided when actions have var
Acked-by: Justin Pettit
--Justin
On Oct 18, 2013, at 1:14 PM, Gurucharan Shetty wrote:
> This commit changes VTEP simulator's flooding logic as follows:
>
> 1. Traffic coming from a VTEP physical port will only be flooded
> to all the physical ports in the same VTEP that belongs to that logi
With mega flow implementation ovs flow can be shared between
multiple CPUs which makes stats updates highly contended
operation. Following patch allocates separate stats for each
CPU to make stats update scalable.
Signed-off-by: Pravin B Shelar
---
v2:
disable bh for local stats read.
v3:
disable
On Fri, Oct 18, 2013 at 1:33 PM, Pravin Shelar wrote:
> On Fri, Oct 18, 2013 at 1:21 PM, Jesse Gross wrote:
>> On Fri, Oct 18, 2013 at 1:06 PM, Pravin B Shelar wrote:
>>> + u8 tcp_flags; /* Union of seen TCP flags. */
>>> +} cacheline_aligned;
>>
>> It probably doesn'
On Fri, Oct 18, 2013 at 1:21 PM, Jesse Gross wrote:
> On Fri, Oct 18, 2013 at 1:06 PM, Pravin B Shelar wrote:
>> diff --git a/datapath/flow.c b/datapath/flow.c
>> index faa4e15..d2ca7fc 100644
>> --- a/datapath/flow.c
>> +++ b/datapath/flow.c
>> +void ovs_flow_stats_get(struct sw_flow *flow, stru
On Fri, Oct 18, 2013 at 1:06 PM, Pravin B Shelar wrote:
> diff --git a/datapath/flow.c b/datapath/flow.c
> index faa4e15..d2ca7fc 100644
> --- a/datapath/flow.c
> +++ b/datapath/flow.c
> +void ovs_flow_stats_get(struct sw_flow *flow, struct sw_flow_stats *res)
> +{
> + int cpu;
> +
> +
This commit changes VTEP simulator's flooding logic as follows:
1. Traffic coming from a VTEP physical port will only be flooded
to all the physical ports in the same VTEP that belongs to that logical
switch and to ONLY one of the devices that have been marked as 'unknown-dst'
2. Traffic coming f
With mega flow implementation ovs flow can be shared between
multiple CPUs which makes stats updates highly contended
operation. Following patch allocates separate stats for each
CPU to make stats update scalable.
Signed-off-by: Pravin B Shelar
---
v2:
disable bh for local stats read.
---
datapa
On Thu, Oct 17, 2013 at 10:07 PM, Ben Pfaff wrote:
> On Thu, Oct 17, 2013 at 10:02:30PM -0700, Ansis Atteka wrote:
>> After introducing commit a5ae88ff8a276e86c842ac102145432662bf711a
>> "ovsdb-doc: generate vswitch.[pic|gv] files only if dot tool is
>> available" Open vSwitch did not build anymor
Thanks, applied to master.
On Fri, Oct 18, 2013 at 11:25:08AM -0700, Alex Wang wrote:
> Looks good to me, thx
>
>
> On Fri, Oct 18, 2013 at 11:20 AM, Ben Pfaff wrote:
>
> > Signed-off-by: Ben Pfaff
> > ---
> > FAQ |2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --gi
Looks good to me, thx
On Fri, Oct 18, 2013 at 11:20 AM, Ben Pfaff wrote:
> Signed-off-by: Ben Pfaff
> ---
> FAQ |2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/FAQ b/FAQ
> index b81ee6c..0a081ff 100644
> --- a/FAQ
> +++ b/FAQ
> @@ -169,7 +169,7 @@ Q: Should userspa
Signed-off-by: Ben Pfaff
---
FAQ |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/FAQ b/FAQ
index b81ee6c..0a081ff 100644
--- a/FAQ
+++ b/FAQ
@@ -169,7 +169,7 @@ Q: Should userspace or kernel be upgraded first to minimize
downtime?
kernel version included in the same re
On Thu, Oct 17, 2013 at 03:12:43PM -0700, Andy Zhou wrote:
> Signed-off-by: Andy Zhou
In show_dpif(), I think it's theoretically possible, at least, for
n_pkts to be 0, in which case the calculation of 'avg' would be a
division by zero. I think we should probably avoid that.
Otherwise:
Acked-by
Thanks, applied to master.
On Fri, Oct 18, 2013 at 10:49:31AM -0700, Justin Pettit wrote:
> Looks great. Thanks!
>
> --Justin
>
>
> On Oct 18, 2013, at 10:29 AM, Ben Pfaff wrote:
>
> > Signed-off-by: Ben Pfaff
> > ---
> > FAQ | 19 +++
> > 1 file changed, 19 insertions(+)
>
Looks great. Thanks!
--Justin
On Oct 18, 2013, at 10:29 AM, Ben Pfaff wrote:
> Signed-off-by: Ben Pfaff
> ---
> FAQ | 19 +++
> 1 file changed, 19 insertions(+)
>
> diff --git a/FAQ b/FAQ
> index 99e6107..b81ee6c 100644
> --- a/FAQ
> +++ b/FAQ
> @@ -198,6 +198,25 @@ A: Tunn
Signed-off-by: Ben Pfaff
---
FAQ | 19 +++
1 file changed, 19 insertions(+)
diff --git a/FAQ b/FAQ
index 99e6107..b81ee6c 100644
--- a/FAQ
+++ b/FAQ
@@ -198,6 +198,25 @@ A: Tunnel virtual ports are not supported, as described in
the
actions. On Linux kernels before 2.6.39
Signed-off-by: Gurucharan Shetty
---
ofproto/ofproto.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/ofproto/ofproto.c b/ofproto/ofproto.c
index f44a7c4..e5b8237 100644
--- a/ofproto/ofproto.c
+++ b/ofproto/ofproto.c
@@ -451,7 +451,7 @@ ofproto_enumerate_names(const char
Signed-off-by: Gurucharan Shetty
---
lib/classifier.h |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/lib/classifier.h b/lib/classifier.h
index 0e39012..8e3bf61 100644
--- a/lib/classifier.h
+++ b/lib/classifier.h
@@ -148,7 +148,7 @@ cls_table_is_catchall(const struct cls_t
On Fri, Oct 18, 2013 at 8:37 AM, Andy Zhou wrote:
> On Thu, Oct 17, 2013 at 4:37 PM, Jesse Gross wrote:
>> One question on the interface - I noticed that the n_masks field is 64
>> bit. I guess this is to provide some kind of regular structure since I
>> hope we never have that many masks...
>
>
I changed NEWS and pushed it to master.
Thanks,
Pravin.
On Wed, Oct 16, 2013 at 3:40 PM, Jesse Gross wrote:
> On Mon, Oct 14, 2013 at 3:26 PM, Pravin B Shelar wrote:
>> diff --git a/FAQ b/FAQ
>> index d36495c..e904d35 100644
>> --- a/FAQ
>> +++ b/FAQ
>> @@ -148,7 +148,7 @@ A: The following tabl
On Fri, Oct 18, 2013 at 12:08 AM, Li, Chen wrote:
> I have changed my test method a little to make the issue more clear:
>
>
>
> ** **
>
> As you can see, my vm has two virtual NIC, tap1 is working under VLAN
> (tag=2002), while tap2 is working under FLAT.
>
> And the bandwidth
On Fri, Oct 18, 2013 at 8:25 AM, Andy Zhou wrote:
> Do we still need a spinlock when we have per-cpu counters?
>
Yes, cmd OVS_FLOW_ATTR_CLEAR needs clear stats. So lock is required
for atomic reset.
>
> On Fri, Oct 18, 2013 at 8:11 AM, Pravin B Shelar wrote:
>>
>> With mega flow implementation
Do we still need a spinlock when we have per-cpu counters?
On Fri, Oct 18, 2013 at 8:11 AM, Pravin B Shelar wrote:
> With mega flow implementation ovs flow can be shared between
> multiple CPUs which makes stats updates highly contended
> operation. Following patch allocates separate stats for
With mega flow implementation ovs flow can be shared between
multiple CPUs which makes stats updates highly contended
operation. Following patch allocates separate stats for each
CPU to make stats update scalable.
Signed-off-by: Pravin B Shelar
---
datapath/datapath.c | 54 ++
Hi, Any thoughts on this subject?
Thanks
Jiri
Tue, Oct 08, 2013 at 04:06:23PM CEST, j...@resnulli.us wrote:
>Hi OVS developers.
>
>I'm a maintainer of Team driver and libteam (userspace counterpart).
>
>Recently I have been looking into ovs userspace and kernel code wrt
>bonding features. Looks
38 matches
Mail list logo
