On Fri, Nov 01, 2013 at 03:23:10PM +0900, YAMAMOTO Takashi wrote:
> > On Tue, Oct 15, 2013 at 05:12:22PM +0900, YAMAMOTO Takashi wrote:
> >> Signed-off-by: YAMAMOTO Takashi
> >> ---
> >> OPENFLOW-1.1+ | 7 +++
> >> 1 file changed, 7 insertions(+)
> >>
> >> diff --git a/OPENFLOW-1.1+ b/OPENFL
> On Fri, Nov 01, 2013 at 03:23:10PM +0900, YAMAMOTO Takashi wrote:
>> > On Tue, Oct 15, 2013 at 05:12:22PM +0900, YAMAMOTO Takashi wrote:
>> >> Signed-off-by: YAMAMOTO Takashi
>> >> ---
>> >> OPENFLOW-1.1+ | 7 +++
>> >> 1 file changed, 7 insertions(+)
>> >>
>> >> diff --git a/OPENFLOW-1.1+
On Fri, Nov 01, 2013 at 04:25:25PM +0900, YAMAMOTO Takashi wrote:
> > On Fri, Nov 01, 2013 at 03:23:10PM +0900, YAMAMOTO Takashi wrote:
> >> > On Tue, Oct 15, 2013 at 05:12:22PM +0900, YAMAMOTO Takashi wrote:
> >> >> Signed-off-by: YAMAMOTO Takashi
> >> >> ---
> >> >> OPENFLOW-1.1+ | 7 +++
>
Is that possible to add ability to bring up ovs interfaces in case it
defined not in /etc/network/interfaces but using source xxx, for
example i have
/etc/network/interfaces:
source /etc/network/interfaces.d/*
and in /etc/network/interfaces.d/XXX i'm configure each interface.
2013/10/31 Guruchar
Thanks for your reply!
On 01/11/2013 12:29, Simon Horman wrote:
On Sat, Oct 26, 2013 at 06:14:28PM +0800, Alexander Wu wrote:
Implement the encode/decode table features msgs function, and
NOTE that we implement the decode functions *_raw, maybe we
should change it the ofpbuf_pull?
Signed-off-b
On 01/11/2013 12:32, Simon Horman wrote:
On Sat, Oct 26, 2013 at 06:12:27PM +0800, Alexander Wu wrote:
Add headers and function prototype for table features.
And modify the limits of mp-table-features msg
Signed-off-by: Alexander Wu
---
lib/ofp-msgs.h |4 +-
lib/ofp-util.h
On 01/11/2013 12:32, Simon Horman wrote:
On Sat, Oct 26, 2013 at 06:15:30PM +0800, Alexander Wu wrote:
Add some functions to init table features(use ofp13_* struct
currently, change it to ofputil later).
Use the encode/decode functions to handle table features request.
Currently we just imple
On Fri, Nov 1, 2013 at 3:13 AM, Vasiliy Tolstov wrote:
> Is that possible to add ability to bring up ovs interfaces in case it
> defined not in /etc/network/interfaces but using source xxx, for
> example i have
> /etc/network/interfaces:
> source /etc/network/interfaces.d/*
>
> and in /etc/networ
Signed-off-by: Ben Pfaff
---
lib/ofp-util.c | 21 +-
tests/learn.at |4 +--
tests/ofp-print.at | 12
tests/ovs-ofctl.at | 48 +++
utilities/ovs-ofctl.8.in | 71 +++-
Jarno pointed out that modify_flows__() didn't really need to check every
instance of the flow separately. After some further investigation I
decided that this was even more of an improvement.
CC: Jarno Rajahalme
Signed-off-by: Ben Pfaff
---
lib/ofp-actions.c | 23 ++-
This function is related to actions to ofp-actions seems like a logical
place for it.
Signed-off-by: Ben Pfaff
---
lib/bundle.c |2 +-
lib/ofp-actions.c | 33 +
lib/ofp-actions.h |1 +
lib/ofp-util.c| 25 -
lib/ofp-util
On Fri, Oct 25, 2013 at 02:25:10PM -0700, Ben Pfaff wrote:
> On Fri, Oct 25, 2013 at 02:04:44PM -0700, Jarno Rajahalme wrote:
> >
> > On Oct 25, 2013, at 1:56 PM, Ben Pfaff wrote:
> >
> > > On Fri, Oct 25, 2013 at 01:44:17PM -0700, Jarno Rajahalme wrote:
> > >>
> > >> On Oct 25, 2013, at 1:32 P
Signed-off-by: Thomas Graf
---
ovsdb/ovsdb-server.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/ovsdb/ovsdb-server.c b/ovsdb/ovsdb-server.c
index ab44b3a..f64122e 100644
--- a/ovsdb/ovsdb-server.c
+++ b/ovsdb/ovsdb-server.c
@@ -983,6 +983,7 @@ ovsdb_server_compact(struct unixctl_conn *con
On Wed, Oct 30, 2013 at 06:17:04PM +0900, Simon Horman wrote:
> With these patches in place it is possible to use all group types
> defined in OpenFlow1.3.
I applied patches 1-4 to master. I haven't had a chance to look at the
rest yet.
___
dev mailing
On Fri, Nov 01, 2013 at 05:44:52PM +0100, Thomas Graf wrote:
> Signed-off-by: Thomas Graf
Thanks, applied to master, branch-2.0, branch-1.11, branch-1.10,
branch-1.9.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
Add a "switch_fault" and "port_fault" column to the appropriate
tables in the VTEP schema.
Signed-off-by: Bruce Davie
---
vtep/vtep.ovsschema | 16
vtep/vtep.xml | 48 ++--
2 files changed, 58 insertions(+), 6 deletions(-)
diff
'struct dp_netdev_flow' is currently being instantiated as 'flow'.
An upcoming commit introduces a classifier to dpif-netdev
which uses 'struct flow' at a few places and that can cause
confusion while reading code.
Signed-off-by: Gurucharan Shetty
---
lib/dpif-netdev.c | 135 +++
Instead of an exact match flow table, we introduce a classifier.
This enables mega-flows in userspace datapath.
Signed-off-by: Gurucharan Shetty
---
lib/dpif-netdev.c | 87 ++
tests/ofproto-dpif.at | 142 +
2 file
Right now, when we want to look at the flows of netdev
datapath, we use the dpif/dump-flows command which provides
it without querying the netdev datapath but rather through the
facets and subfacets.
This commit adds the ability to directly query the netdev
datapath through the dpif interface.
Si
Signed-off-by: Gurucharan Shetty
---
lib/dpif-netdev.c | 13 ++---
1 file changed, 10 insertions(+), 3 deletions(-)
diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c
index ea8a6d3..ae1fecd 100644
--- a/lib/dpif-netdev.c
+++ b/lib/dpif-netdev.c
@@ -987,6 +987,7 @@ struct dp_netdev_flow
Signed-off-by: Gurucharan Shetty
---
ofproto/ofproto-dpif.c | 26 ++
vswitchd/ovs-vswitchd.8.in |7 ++-
2 files changed, 32 insertions(+), 1 deletion(-)
diff --git a/ofproto/ofproto-dpif.c b/ofproto/ofproto-dpif.c
index 1a53766..2d6ea61 100644
--- a/ofproto/
Currently, with 'ovs-dpctl show', we call show_dpif().
Move this functionality to lib/dpif.c so that it can
be used by an upcoming commit.
Signed-off-by: Gurucharan Shetty
---
lib/dpif.c| 136 +++
lib/dpif.h|1 +
utilities/
On Fri, Nov 01, 2013 at 02:15:12AM -0700, Gurucharan Shetty wrote:
> 'struct dp_netdev_flow' is currently being instantiated as 'flow'.
> An upcoming commit introduces a classifier to dpif-netdev
> which uses 'struct flow' at a few places and that can cause
> confusion while reading code.
>
> Sign
When dealing with a large number of ports, one of the performance
bottlenecks is that we loop through all netdevs in the main loop. Miimon
is a contributor to this, executing even if it is not enabled on any
devices.
This patch introduces a counter for the number of netdevs with miimon
configured.
On Fri, Nov 01, 2013 at 02:15:13AM -0700, Gurucharan Shetty wrote:
> Instead of an exact match flow table, we introduce a classifier.
> This enables mega-flows in userspace datapath.
>
> Signed-off-by: Gurucharan Shetty
Clang says:
../lib/dpif-netdev.c:638:5: error: calling function 'classi
On Fri, Nov 01, 2013 at 01:06:39PM -0700, Joe Stringer wrote:
> When dealing with a large number of ports, one of the performance
> bottlenecks is that we loop through all netdevs in the main loop. Miimon
> is a contributor to this, executing even if it is not enabled on any
> devices.
>
> This pa
On Fri, Nov 01, 2013 at 02:15:14AM -0700, Gurucharan Shetty wrote:
> Signed-off-by: Gurucharan Shetty
Any reason this isn't part of patch 2? It seems to be needed.
Why is the in_port mask hard-coded to UINT32_MAX? I don't remember
anything in patch 2 that always un-wildcarded the in_port. (Ma
On Fri, Nov 01, 2013 at 02:15:15AM -0700, Gurucharan Shetty wrote:
> Right now, when we want to look at the flows of netdev
> datapath, we use the dpif/dump-flows command which provides
> it without querying the netdev datapath but rather through the
> facets and subfacets.
>
> This commit adds th
On Fri, Nov 01, 2013 at 02:15:16AM -0700, Gurucharan Shetty wrote:
> Currently, with 'ovs-dpctl show', we call show_dpif().
> Move this functionality to lib/dpif.c so that it can
> be used by an upcoming commit.
>
> Signed-off-by: Gurucharan Shetty
Acked-by: Ben Pfaff
__
On Fri, Nov 01, 2013 at 02:15:17AM -0700, Gurucharan Shetty wrote:
> Signed-off-by: Gurucharan Shetty
Again this looks useful but why not put it in lib/dpif.c?
Thanks,
Ben.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listin
On Fri, Nov 01, 2013 at 10:53:23AM -0700, Bruce Davie wrote:
> Add a "switch_fault" and "port_fault" column to the appropriate
> tables in the VTEP schema.
>
> Signed-off-by: Bruce Davie
Applied, thanks.
___
dev mailing list
dev@openvswitch.org
http://
On Thu, Oct 24, 2013 at 01:19:25PM -0700, Jarno Rajahalme wrote:
> OpenFlow 1.1 set vlan actions only modify existing vlan
> headers, while OF 1.0 actions push a new vlan header if one
> does not exist already.
>
> Signed-off-by: Jarno Rajahalme
Applied. I folded in a couple of minor improvemen
Nothing about the existing code in this function cares whether Goto-Table
is the last action.
Signed-off-by: Ben Pfaff
---
ofproto/ofproto-dpif-xlate.c |1 -
1 file changed, 1 deletion(-)
diff --git a/ofproto/ofproto-dpif-xlate.c b/ofproto/ofproto-dpif-xlate.c
index f19dbf1..f5bc12c 100644
On Thu, Oct 24, 2013 at 01:19:26PM -0700, Jarno Rajahalme wrote:
> Store the original action code with the strip vlan action,
> so that it can be printed back properly.
>
> Signed-off-by: Jarno Rajahalme
Applied, thanks!
___
dev mailing list
dev@openvs
On Thu, Oct 24, 2013 at 01:19:27PM -0700, Jarno Rajahalme wrote:
> mf_from_id accesses a static table, so the compiler should be able to
> completely optimize it away.
>
> Also use OVS_PACKED_ENUM to waste less space.
>
> Signed-off-by: Jarno Rajahalme
Applied, thanks.
I added a "meta-flow: "
LGTM,
On Fri, Nov 1, 2013 at 2:08 PM, Ben Pfaff wrote:
> Nothing about the existing code in this function cares whether Goto-Table
> is the last action.
>
> Signed-off-by: Ben Pfaff
> ---
> ofproto/ofproto-dpif-xlate.c |1 -
> 1 file changed, 1 deletion(-)
>
> diff --git a/ofproto/ofproto
Signed-off-by: Alexandru Copot
---
ovsdb/ovsdb-tool.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/ovsdb/ovsdb-tool.c b/ovsdb/ovsdb-tool.c
index 11e61e6..5e2b71b 100644
--- a/ovsdb/ovsdb-tool.c
+++ b/ovsdb/ovsdb-tool.c
@@ -524,7 +524,7 @@ do_show_log(int argc, char *argv[])
Hopefully I got the right thread here ("FAQ: Elaborate further on how
one drops packets with OpenFlow.").
Acked-by: Joe Stringer
On Wed, Oct 9, 2013 at 3:22 PM, Ben Pfaff wrote:
>
> Signed-off-by: Ben Pfaff
> ---
> FAQ | 11 ---
> 1 file changed, 8 insertions(+), 3 deletions(-)
>
>
On Thu, Oct 24, 2013 at 01:19:28PM -0700, Jarno Rajahalme wrote:
> Adds OXM inspired aliases for match fields that don't have them
> already ("ip_proto", "ip_ecn", "ip_dscp", and "tunnel_id").
>
> "ip_dscp" replaces the earlier undocumented "nw_tos_shifted",
> and takes the DSCP value (0-63), whic
On 1 November 2013 13:31, Ben Pfaff wrote:
> On Fri, Nov 01, 2013 at 01:06:39PM -0700, Joe Stringer wrote:
> > When dealing with a large number of ports, one of the performance
> > bottlenecks is that we loop through all netdevs in the main loop. Miimon
> > is a contributor to this, executing eve
Thanks.
On Fri, Nov 01, 2013 at 02:17:11PM -0700, Alex Wang wrote:
> LGTM,
>
>
> On Fri, Nov 1, 2013 at 2:08 PM, Ben Pfaff wrote:
>
> > Nothing about the existing code in this function cares whether Goto-Table
> > is the last action.
> >
> > Signed-off-by: Ben Pfaff
> > ---
> > ofproto/ofpro
On Fri, Nov 01, 2013 at 11:20:56PM +0200, Alexandru Copot wrote:
> Signed-off-by: Alexandru Copot
Applied, thanks.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
Acked-by: Joe Stringer
On Fri, Oct 11, 2013 at 4:53 PM, Ben Pfaff wrote:
> Signed-off-by: Ben Pfaff
> ---
> lib/daemon.c |2 ++
> lib/daemon.man |7 ---
> 2 files changed, 6 insertions(+), 3 deletions(-)
>
> diff --git a/lib/daemon.c b/lib/daemon.c
> index 1c9ebe2..54641d0 100644
On Thu, Oct 24, 2013 at 01:19:29PM -0700, Jarno Rajahalme wrote:
> Signed-off-by: Jarno Rajahalme
I'm going to fold this in:
diff --git a/lib/ofp-actions.c b/lib/ofp-actions.c
index a9d4340..9f31449 100644
--- a/lib/ofp-actions.c
+++ b/lib/ofp-actions.c
@@ -793,7 +793,7 @@ set_field_from_openflo
My name is Mark Carney, I am the new Governor of Bank Of England, There is the
sum of 22,400,000.00 Pounds in my Bank, There were no beneficiaries stated
concerning these funds which means no one would ever come forward to claim it.
Please write me back if interested for further explanation. Rep
v2: Properly finish hashes, more flexible configuration.
Signed-off-by: Jarno Rajahalme
---
lib/classifier.c | 174 +++---
lib/classifier.h | 12 ++-
lib/flow.c | 163 ++-
lib/fl
This series implements two techniques for better wildcarding:
1. Staged sub-table matching: Each sub-table is matched in segments,
starting from metadata and lower protocol layer fields, progressing
towards higher layers only as needed. That is, if we can determine
that there is no match
There is a bit more bookkeeping, and preliminary measurements did
not show significant benefit. Results may differ with larger
flow tables, though.
Signed-off-by: Jarno Rajahalme
---
lib/classifier.c | 156 +-
1 file changed, 84 insertions(+)
Add a prefix tree (trie) structure for tracking the used
IP address space, enabling skipping classifier tables
containing longer masks than necessary for the given
address. This enables more wildcarding for megaflows in
parts of the address space without host routes.
In fact, the trie lookups resu
Fair enough ;-)
Jarno
On Nov 1, 2013, at 3:02 PM, Ben Pfaff wrote:
> On Thu, Oct 24, 2013 at 01:19:29PM -0700, Jarno Rajahalme wrote:
>> Signed-off-by: Jarno Rajahalme
>
> I'm going to fold this in:
>
> diff --git a/lib/ofp-actions.c b/lib/ofp-actions.c
> index a9d4340..9f31449 100644
> --
Acked-by: Jarno Rajahalme
On Nov 1, 2013, at 9:43 AM, Ben Pfaff wrote:
> This function is related to actions to ofp-actions seems like a logical
> place for it.
>
> Signed-off-by: Ben Pfaff
> ---
> lib/bundle.c |2 +-
> lib/ofp-actions.c | 33 +
> lib/
This commit improves the upcall dispatching fairness by introduing
a 2-stage scheme. At the first stage, the dispatcher thread will
try receiving upcalls from dpif and hashing the upcalls to the local
queues based on the L2 header information. When either there is no
upcall to receive or one of t
Hey Ethan,
This is the implementation of the 2-stage scheme.
My experiment shows that it does help reduce both the loss rate and
handling latency. I'll format my test result and attach it to this thread
later.
Thanks,
___
dev mailing list
dev@openvswi
On Nov 1, 2013, at 9:43 AM, Ben Pfaff wrote:
> Signed-off-by: Ben Pfaff
> ---
> lib/ofp-util.c | 21 +-
> tests/learn.at |4 +--
> tests/ofp-print.at | 12
> tests/ovs-ofctl.at | 48 +++
> utilities/ovs-ofctl
On Fri, Nov 01, 2013 at 03:22:34PM -0700, Jarno Rajahalme wrote:
> Fair enough ;-)
Thanks.
I'm folding in the following additional changes, and then I'll apply
this to master.
The only really notable part of the changes that I made is that I'm
feeling timid about silently adding and removing the
On Nov 1, 2013, at 9:43 AM, Ben Pfaff wrote:
> Jarno pointed out that modify_flows__() didn't really need to check every
> instance of the flow separately. After some further investigation I
> decided that this was even more of an improvement.
>
Indeed! Check the actions right after they are
This needs a rebase.
The deletion of the content in 'manpages.mk' is surprising; the file
states that it is automatically generated, the commit message doesn't
mention it, and the file isn't actually deleted, only emptied. Was
this intentional?
There seems to be another reference to "openvswitch-
On Thu, Oct 24, 2013 at 01:19:30PM -0700, Jarno Rajahalme wrote:
> Output set field actions as standard OF1.0/1.1 set actions or to
> reg_load instructions, when a compatible set action(s) do not exist.
>
> Signed-off-by: Jarno Rajahalme
Thanks.
I broke set_field_to_openflow() into a couple of
On Nov 1, 2013, at 4:04 PM, Ben Pfaff wrote:
> On Thu, Oct 24, 2013 at 01:19:30PM -0700, Jarno Rajahalme wrote:
>> Output set field actions as standard OF1.0/1.1 set actions or to
>> reg_load instructions, when a compatible set action(s) do not exist.
>>
>> Signed-off-by: Jarno Rajahalme
>
>
On Thu, Oct 24, 2013 at 01:19:31PM -0700, Jarno Rajahalme wrote:
> This makes later changes simpler.
>
> Signed-off-by: Jarno Rajahalme
Thanks! Applied to master.
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Thu, Oct 24, 2013 at 01:19:32PM -0700, Jarno Rajahalme wrote:
> Signed-off-by: Jarno Rajahalme
Applied, thanks!
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
This looks good in principle.
Test 706 (ofproto-dpif - too many output actions) fails on this, as it
expects the old VLOG_ERR in the logs.
There is also an error from ofproto_dpif_xlate in this case,
complaining about resubmits yielding >64kB of actions. I assume this
is intended to remain as a h
When dealing with a large number of ports, one of the performance
bottlenecks is that we loop through all netdevs in the main loop. Miimon
is a contributor to this, checking all devices even if it has never been
enabled.
This patch introduces a counter for the number of netdevs with miimon
configu
Acked-by: Joe Stringer
On Thu, Oct 17, 2013 at 3:29 PM, Ben Pfaff wrote:
> The ARP headers have been acceptable as NXAST_REG_MOVE destinations since
> commit f6c8a6b163 (Add software switch support for modifying ARP headers
> in OpenFlow.)
>
> Reported-by: Anupam Chanda
> Signed-off-by: Ben Pfa
On Fri, Nov 01, 2013 at 03:42:30PM -0700, Jarno Rajahalme wrote:
>
> On Nov 1, 2013, at 9:43 AM, Ben Pfaff wrote:
>
> > Signed-off-by: Ben Pfaff
> > ---
> > lib/ofp-util.c | 21 +-
> > tests/learn.at |4 +--
> > tests/ofp-print.at | 12
> > te
On Fri, Nov 01, 2013 at 03:58:52PM -0700, Jarno Rajahalme wrote:
>
> On Nov 1, 2013, at 9:43 AM, Ben Pfaff wrote:
>
> > Jarno pointed out that modify_flows__() didn't really need to check every
> > instance of the flow separately. After some further investigation I
> > decided that this was eve
On Thu, Oct 24, 2013 at 01:19:33PM -0700, Jarno Rajahalme wrote:
> Signed-off-by: Jarno Rajahalme
Applied, thanks!
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Thu, Oct 24, 2013 at 01:19:34PM -0700, Jarno Rajahalme wrote:
> Signed-off-by: Jarno Rajahalme
Applied, thanks!
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
On Thu, Oct 31, 2013 at 04:16:28PM +0900, Simon Horman wrote:
> This series implements MPLS actions and matches based on work by
> Ravi K, Leo Alterman, Yamahata-san and Joe Stringer.
Would you mind rebasing this one more time?
___
dev mailing list
dev@o
2013/11/1 Gurucharan Shetty :
> It is not possible with the current code. It only looks at
> /etc/network/interfaces
But if for example cat all files to buffer in case of source and
operates not on interfaces but on buffered contents of all files?
--
Vasiliy Tolstov,
e-mail: v.tols...@selfip.ru
On Wed, Oct 30, 2013 at 06:17:09PM +0900, Simon Horman wrote:
> This is a first step towards implementing the dpif side of groups.
>
> In order to be useful the action translation code needs
> to be taught about groups.
>
> Signed-off-by: Simon Horman
Applied, thanks!
__
On Wed, Oct 30, 2013 at 06:17:10PM +0900, Simon Horman wrote:
> Require liveness for fast-failover groups as it is mandated by OpenFlow1.3.
> Allow livness for select groups which is in keeping with OpenFlow1.3.
> Disallow liveness it for other group types.
>
> Signed-off-by: Simon Horman
Applie
72 matches
Mail list logo
