On Wed, Mar 27, 2013 at 11:02:21PM -0700, Neil Mckee wrote:
> This patch adds an sFlow test to the test suite (in branch 1.10). To
> make that work properly I added netdev_dummy_get_ifindex() so that a
> dummy netdev can return a dummy ifindex when asked. Is there anywhere
> in OVS that assumes t
This patch adds an sFlow test to the test suite (in branch 1.10). To make that
work properly I added netdev_dummy_get_ifindex() so that a dummy netdev can
return a dummy ifindex when asked. Is there anywhere in OVS that assumes that
a netdev_dummy cannot make up a dummy ifindex? If so, I gue
On Mon, Mar 25, 2013 at 08:18:02AM +0900, Simon Horman wrote:
> Recirculation is a technique to allow a frame to re-enter
> frame processing. This is intended to be used after actions
> have been applied to the frame with modify the frame in
> some way that makes it possible for richer processing t
On Thu, Mar 21, 2013 at 04:50:03PM +0900, Simon Horman wrote:
> Allow datapath to recognize and extract MPLS labels into flow keys
> and execute actions which push, pop, and set labels on packets.
>
> Based heavily on work by Leo Alterman and Ravi K.
>
> Cc: Ravi K
> Cc: Leo Alterman
> Reviewed
On Thu, Mar 28, 2013 at 01:21:08AM -0400, David Miller wrote:
> From: Simon Horman
> Date: Thu, 28 Mar 2013 13:38:25 +0900
>
> > Add a new constant ETH_P_802_3_MIN, the minimum ethernet type for
> > an 802.3 frame. Frames with a lower value in the ethernet type field
> > are Ethernet II.
> >
> >
From: Simon Horman
Date: Thu, 28 Mar 2013 13:38:25 +0900
> Add a new constant ETH_P_802_3_MIN, the minimum ethernet type for
> an 802.3 frame. Frames with a lower value in the ethernet type field
> are Ethernet II.
>
> Also update all the users of this value that David Miller and
> I could find
Add a new constant ETH_P_802_3_MIN, the minimum ethernet type for
an 802.3 frame. Frames with a lower value in the ethernet type field
are Ethernet II.
Also update all the users of this value that David Miller and
I could find to use the new constant.
Also correct a bug in util.c. The comparison
On Wed, Mar 27, 2013 at 06:56:02PM -0700, Andy Zhou wrote:
> Thanks Ben for explaining. The patch looks good.
Thanks.
> If we skip 0s in flow_hash() as well, then miniflow_hash() can be made to
> produce the same value as flow_hash(), if at all useful.
I agree that it is possible. I do not kno
This sounds attractive. Let's talk about how it should look like in more
detail.
On Wed, Mar 27, 2013 at 4:18 PM, Ben Pfaff wrote:
> On Wed, Mar 27, 2013 at 04:01:57PM -0700, Ethan Jackson wrote:
> > I had a question about this. Did we consider simply using coverage
> > counters? We could ext
Thanks Ben for explaining. The patch looks good.
If we skip 0s in flow_hash() as well, then miniflow_hash() can be made to
produce the same value as flow_hash(), if at all useful.
Andy
On Wed, Mar 27, 2013 at 5:14 PM, Ben Pfaff wrote:
> On Wed, Mar 27, 2013 at 12:14:52PM -0700, Andy Zhou wro
On Wed, Mar 27, 2013 at 12:14:52PM -0700, Andy Zhou wrote:
> It would be nice to add a comment about what 'n' is.
>
> > +static void
> > +miniflow_init__(struct miniflow *dst, const struct flow *src, int n)
I added this comment:
/* Completes an initialization of 'dst' as a miniflow copy of 'src'
On Wed, Mar 27, 2013 at 04:01:57PM -0700, Ethan Jackson wrote:
> I had a question about this. Did we consider simply using coverage
> counters? We could extend them to report a rate. This would be
> useful more generally.
Andy and I talked about this briefly this afternoon. I think that it
wou
On Thu, Mar 21, 2013 at 03:00:29PM -0700, Andy Zhou wrote:
> When error is encountered with add-flows command, this patch
> adds file name and line number in addition to the parser
> error message to the output.
>
> The file name and line number will not be added to
> the output of "ovs-ofctl add-
On Wed, Mar 27, 2013 at 3:38 PM, Ben Pfaff wrote:
> On Wed, Mar 27, 2013 at 02:30:05PM -0700, Gurucharan Shetty wrote:
>> Currently, when we stop a daemon, we first send it SIGTERM.
>> If SIGTERM did not work within ~5 seconds, we send a SIGKILL.
>> After sending SIGKILL, we wait only for 4 second
I had a question about this. Did we consider simply using coverage
counters? We could extend them to report a rate. This would be
useful more generally.
Ethan
On Wed, Mar 27, 2013 at 3:59 PM, Andy Zhou wrote:
> This is fine. Thanks
>
>
> On Wednesday, March 27, 2013, Ben Pfaff wrote:
>>
>> On
Thanks, applied to master.
On Wed, Mar 27, 2013 at 03:59:03PM -0700, Andy Zhou wrote:
> This is fine. Thanks
>
> On Wednesday, March 27, 2013, Ben Pfaff wrote:
>
> > On Mon, Mar 25, 2013 at 07:49:13PM -0700, Andy Zhou wrote:
> > > This patch adds more flow related stats to the output of
> > > "o
This is fine. Thanks
On Wednesday, March 27, 2013, Ben Pfaff wrote:
> On Mon, Mar 25, 2013 at 07:49:13PM -0700, Andy Zhou wrote:
> > This patch adds more flow related stats to the output of
> > "ovs-appctl dpif/show". Specifically, the follow information
> > are added per ofproto:
> >
> > - Max
On Mon, Mar 25, 2013 at 07:49:13PM -0700, Andy Zhou wrote:
> This patch adds more flow related stats to the output of
> "ovs-appctl dpif/show". Specifically, the follow information
> are added per ofproto:
>
> - Max flow table size
> - Average flow table size
> - Average flow table add rate
> - A
On Wed, Mar 27, 2013 at 02:30:05PM -0700, Gurucharan Shetty wrote:
> Currently, when we stop a daemon, we first send it SIGTERM.
> If SIGTERM did not work within ~5 seconds, we send a SIGKILL.
> After sending SIGKILL, we wait only for 4 seconds, before giving
> up.
>
> If the system is exteremely
Thanks, I pushed this to master.
On Wed, Mar 27, 2013 at 02:44:56PM -0700, Justin Pettit wrote:
> Looks good.
>
> --Justin
>
>
> On Mar 27, 2013, at 2:41 PM, Ben Pfaff wrote:
>
> > Signed-off-by: Ben Pfaff
> > ---
> > ovsdb/ovsdb.h |6 +-
> > 1 files changed, 1 insertions(+), 5 deleti
Currently, when we stop a daemon, we first send it SIGTERM.
If SIGTERM did not work within ~5 seconds, we send a SIGKILL.
After sending SIGKILL, we wait only for 4 seconds, before giving
up.
If the system is exteremely busy, there is a chance that a
process is not killed by the kernel within 4 sec
Looks good.
--Justin
On Mar 27, 2013, at 2:41 PM, Ben Pfaff wrote:
> Signed-off-by: Ben Pfaff
> ---
> ovsdb/ovsdb.h |6 +-
> 1 files changed, 1 insertions(+), 5 deletions(-)
>
> diff --git a/ovsdb/ovsdb.h b/ovsdb/ovsdb.h
> index 6e4ff79..e976d3b 100644
> --- a/ovsdb/ovsdb.h
> +++ b/ov
These will be used for the first time in an upcoming commit.
Signed-off-by: Ben Pfaff
---
lib/ovsdb-data.c | 93 +++-
lib/ovsdb-data.h |6 +++-
tests/test-ovsdb.c | 14 ++--
3 files changed, 108 insertions(+), 5 deletions(-)
diff -
Signed-off-by: Ben Pfaff
---
tests/ovsdb-server.at | 85 +
1 files changed, 85 insertions(+), 0 deletions(-)
diff --git a/tests/ovsdb-server.at b/tests/ovsdb-server.at
index 62eae38..31f8af3 100644
--- a/tests/ovsdb-server.at
+++ b/tests/ovsdb-se
This series of patches fixes a possible memory exhaustion problem
observed in ovsdb-server in an intentional stress test. The first
three patches fix the problem:
json: New function json_serialized_length().
ovsdb-data: New functions for predicting serialized length of data.
jsonrpc-server:
Users may find this feature useful, but the reason to add this feature is
to allow a test to use it in an upcoming commit.
Signed-off-by: Ben Pfaff
---
NEWS|3 +
ovsdb/ovsdb-client.1.in | 10 ++
ovsdb/ovsdb-client.c| 279
Consider this situation:
* OVSDB client A executes transactions very quickly for a long time.
* OVSDB client B monitors the tables that A modifies, but (either
because B is connected over a slow network, or because B is slow to
process updates) cannot keep up.
In this situati
This will be used for the first time in an upcoming commit.
Signed-off-by: Ben Pfaff
---
lib/json.c| 125 -
lib/json.h|4 +-
tests/test-json.c |3 +-
3 files changed, 129 insertions(+), 3 deletions(-)
diff --git a/lib/
Signed-off-by: Ben Pfaff
---
ovsdb/ovsdb.h |6 +-
1 files changed, 1 insertions(+), 5 deletions(-)
diff --git a/ovsdb/ovsdb.h b/ovsdb/ovsdb.h
index 6e4ff79..e976d3b 100644
--- a/ovsdb/ovsdb.h
+++ b/ovsdb/ovsdb.h
@@ -1,4 +1,4 @@
-/* Copyright (c) 2009, 2010, 2011, 2012 Nicira, Inc.
+/* Co
looks good. A question: Why is full mask of 64-bit metadata required to be
considered for partition? It seems allow partial mask can be more general.
On Tue, Mar 26, 2013 at 9:32 PM, Ben Pfaff wrote:
> We have a controller that puts many rules with different metadata values
> into the flow tab
looks good
On Tue, Mar 26, 2013 at 9:32 PM, Ben Pfaff wrote:
> These seem slightly nicer than e.g. htons(UINT16_MAX).
>
> Signed-off-by: Ben Pfaff
> ---
> include/openvswitch/types.h |6 +-
> lib/match.c | 30 +++---
> lib/meta-flow.c
-- Forwarded message --
From: Andy Zhou
Date: Wed, Mar 27, 2013 at 12:14 PM
Subject: Re: [ovs-dev] [PATCH 1/3] match: New function
minimatch_matches_flow().
To: Ben Pfaff
Looks good. A few minor comments in-line:
On Tue, Mar 26, 2013 at 9:32 PM, Ben Pfaff wrote:
> Signed-off
On Tue, Mar 26, 2013 at 10:09:35PM -0700, Ethan Jackson wrote:
> In the TCP_CRR benchmark, ovs-vswitchd spends so much time in
> update_stats() that it has a significant impact on flow setup
> performance. Further work is needed in this area, but for now,
> simply rate limiting facet_learn() has a
On Tue, Mar 26, 2013 at 10:09:34PM -0700, Ethan Jackson wrote:
> With complex flow tables, facet_check_consistency() can be
> expensive enough to show up in flow setup performance benchmarks.
> In my testing this patch gives us a roughly 10% improvement in
> TCP_CRR and ovs-benchmark.
>
> Signed-o
On Wed, Mar 27, 2013 at 5:41 AM, Hong Zhiguo wrote:
> table->count is uint32_t
>
> Signed-off-by: Hong Zhiguo
Applied, thanks.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
table->count is uint32_t
Signed-off-by: Hong Zhiguo
---
net/openvswitch/flow.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/openvswitch/flow.c b/net/openvswitch/flow.c
index fe0e421..67a2b78 100644
--- a/net/openvswitch/flow.c
+++ b/net/openvswitch/flow.c
@@ -795,9
36 matches
Mail list logo
