David Miller wrote:
>
> I would like to see some discussion wrt. Jamal's feedback, which is that
> a lot of the side-band functionality added by this code is either 1) already
> doable with packet scheduler actions or 2) should be implemented there.
I mostly agree with Jamal. As far as the conc
* John Fastabend (john.r.fastab...@intel.com) wrote:
> He is pushing and popping entire tags off 802.1Q for now but
> you can easily imagine MPLS tags and all sorts of other things
> people will _need_.
Right, already doing some generic encap/decap for tunnelling.
> Do we want tc and likely the s
* Stephen Hemminger (shemmin...@vyatta.com) wrote:
> Maybe someone with more insight than me can explain the relationship
> between Openflow and Open vSwitch. It maybe that the portability
> of Openflow makes the old qdisc, classifiers to use/implement.
I'm sure I can't answer the last bit as well
Looks good.
Ethan
On Tue, Nov 15, 2011 at 17:17, Ben Pfaff wrote:
> An upcoming commit will add a new user.
> ---
> lib/dpif-netdev.c | 20 +---
> lib/packets.c | 21 +
> lib/packets.h | 1 +
> 3 files changed, 23 insertions(+), 19 deletions(-)
Looks good.
Ethan
On Tue, Nov 15, 2011 at 17:17, Ben Pfaff wrote:
> ---
> lib/vlan-bitmap.c | 26 ++
> lib/vlan-bitmap.h | 3 +++
> 2 files changed, 21 insertions(+), 8 deletions(-)
>
> diff --git a/lib/vlan-bitmap.c b/lib/vlan-bitmap.c
> index 94059c7..4c5c1eb 1006
Looks good.
Ethan
On Tue, Nov 15, 2011 at 17:17, Ben Pfaff wrote:
> ---
> lib/automake.mk | 2 +
> lib/netdev-linux.c | 8 ++
> lib/netdev-linux.h | 1 +
> lib/vlandev.c | 213
>
> lib/vlandev.h | 52 +
>
It didn't change. Thanks.
On Tue, Nov 22, 2011 at 07:08:29PM -0800, Ethan Jackson wrote:
> Looks good.
>
> I'm assuming the code of netdev_linux_ethool_set_flag() didn't change
> I didn't read it.
>
> Ethan
>
> On Tue, Nov 15, 2011 at 17:17, Ben Pfaff wrote:
> > ---
> > ?lib/netdev-linux.c |
Looks good.
I'm assuming the code of netdev_linux_ethool_set_flag() didn't change
I didn't read it.
Ethan
On Tue, Nov 15, 2011 at 17:17, Ben Pfaff wrote:
> ---
> lib/netdev-linux.c | 97 ++-
> 1 files changed, 49 insertions(+), 48 deletions(-)
Looks good.
Ethan
On Tue, Nov 15, 2011 at 17:17, Ben Pfaff wrote:
> It has no remaining users.
> ---
> lib/netdev-dummy.c | 1 -
> lib/netdev-linux.c | 56
> -
> lib/netdev-provider.h | 11 -
> lib/netdev-vport.c | 1 -
On Tue, 2011-11-22 at 18:30 -0800, John Fastabend wrote:
> He is pushing and popping entire tags off 802.1Q for now but
> you can easily imagine MPLS tags and all sorts of other things
> people will _need_.
Lots of packet munging already happening with actions.
We can pedit/nat/iptables/checksum
Looks good.
Ethan
On Tue, Nov 15, 2011 at 17:17, Ben Pfaff wrote:
> We introduced this special case before the XenServer integration was
> complete. At that point, we were using VLAN devices on XenServer, with a
> separate bridge for each VLAN, so we needed this special case. But no
> version
On 11/22/2011 5:45 PM, Jamal Hadi Salim wrote:
> On Tue, 2011-11-22 at 15:11 -0800, Jesse Gross wrote:
>
>
>> As you mention, one of the biggest benefits of Open vSwitch is how
>> simple the kernel portions are (it's less than 6000 lines).
>
> I said that was the reason _you_ were using to jus
On Tue, 2011-11-22 at 15:11 -0800, Jesse Gross wrote:
> As you mention, one of the biggest benefits of Open vSwitch is how
> simple the kernel portions are (it's less than 6000 lines).
I said that was the reason _you_ were using to justify things
and i argue that is not accurate.
You will be a
From: Jesse Gross
Date: Tue, 22 Nov 2011 17:34:57 -0800
> On Tue, Nov 22, 2011 at 3:19 PM, David Miller wrote:
>> From: Jesse Gross
>> Date: Tue, 22 Nov 2011 15:11:33 -0800
>>
>>> As you mention, one of the biggest benefits of Open vSwitch is how
>>> simple the kernel portions are (it's less th
On Tue, Nov 22, 2011 at 3:19 PM, David Miller wrote:
> From: Jesse Gross
> Date: Tue, 22 Nov 2011 15:11:33 -0800
>
>> As you mention, one of the biggest benefits of Open vSwitch is how
>> simple the kernel portions are (it's less than 6000 lines). It's
>> existed as an out-of-tree project for se
The classes are never modified and may be declared constant.
---
Compile tested only
$ gcc --version
gcc (Debian 4.6.1-4) 4.6.1
Copyright (C) 2011 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or F
> Oh, except that could you make the subject describe the change instead
> of the bug, e.g. "ofproto-dpif: Properly update tos and ttl fields."
Changed in my local branch.
Ethan
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/lis
On Tue, Nov 22, 2011 at 04:56:02PM -0800, Ben Pfaff wrote:
> On Tue, Nov 22, 2011 at 04:29:34PM -0800, Ethan Jackson wrote:
> > ofproto-dpif failed to update the base flow's tos and ttl fields
> > when preparing for an output action. This could cause redundant
> > updates of those fields in the da
On Tue, Nov 22, 2011 at 04:29:34PM -0800, Ethan Jackson wrote:
> ofproto-dpif failed to update the base flow's tos and ttl fields
> when preparing for an output action. This could cause redundant
> updates of those fields in the datapath. A future patch adds a
> test which could have caught the i
On Tue, Nov 15, 2011 at 05:19:16PM -0800, Ben Pfaff wrote:
> On Tue, Nov 08, 2011 at 11:07:58AM -0800, Jesse Gross wrote:
> > On Mon, Nov 7, 2011 at 4:08 PM, Ben Pfaff wrote:
> > > On Mon, Nov 07, 2011 at 03:54:40PM -0800, Jesse Gross wrote:
> > >> On Mon, Nov 7, 2011 at 1:41 PM, Ben Pfaff wrote:
The dscp column of the queue table instructs Open vSwitch to mark
all traffic egressing the queue with the given DSCP bits in its tos
field.
Bug #7046.
---
ofproto/ofproto-dpif.c | 101
ofproto/ofproto-provider.h |9
ofproto/ofproto.c
Before executing an output action, ofproto-dpif must commit the
changes it's made to the flow so they are reflected in the
packet. This code has been unnecessarily complex. This patch
attempts to simplify the code in the following ways.
- Commit in fewer places.
In an attempt to provide some opt
ofproto-dpif failed to update the base flow's tos and ttl fields
when preparing for an output action. This could cause redundant
updates of those fields in the datapath. A future patch adds a
test which could have caught the issue for the tos bits.
---
ofproto/ofproto-dpif.c |4 ++--
1 files
The key to getting good performance on the netperf CRR test seems to be to
handle the first packet of each new flow as quickly as possible. Until
now, we've only had one opportunity to do that on each trip through the
main poll loop. One way to improve would be to make that poll loop
circulate mo
Looks good.
Ethan
On Tue, Nov 15, 2011 at 17:17, Ben Pfaff wrote:
> NetFlow active timeouts were only mixed in with flow expiration for
> convenience: both processes need to iterate all the facets. But
> an upcoming commit will change flow expiration to work in terms of
> a new "subfacet" entit
From: Jesse Gross
Date: Tue, 22 Nov 2011 15:11:33 -0800
> As you mention, one of the biggest benefits of Open vSwitch is how
> simple the kernel portions are (it's less than 6000 lines). It's
> existed as an out-of-tree project for several years now so it's
> actually fairly mature already and u
Maybe someone with more insight than me can explain the relationship
between Openflow and Open vSwitch. It maybe that the portability
of Openflow makes the old qdisc, classifiers to use/implement.
___
dev mailing list
dev@openvswitch.org
http://openvswitc
The key to getting good performance on the netperf CRR test seems to be to
handle the first packet of each new flow as quickly as possible. Until
now, we've only had one opportunity to do that on each trip through the
main poll loop. One way to improve would be to make that poll loop
circulate mo
This yields a 27% improvement in netperf CRR results in my tests
versus the previous commit, which is a 52% improvement versus
the baseline from just before the poll_fd_woke() optimization was
removed.
---
ofproto/ofproto-dpif.c | 95 +---
1 files chan
epoll appears to be much more efficient than poll() at least for
static file descriptor sets. I can't otherwise explain why this
patch increases netperf CRR performance by 20% above the previous
commit, which is also about a 19% overall improvement versus
the baseline from before the poll_fd_woke(
Using poll() internally in dpif_linux_recv(), instead of relying
on the results of the main loop poll() call, brings netperf CRR
performance back within 1% of par versus the code base before the
poll_fd_woke() optimizations were introduced. It also increases
the ovs-benchmark results by about 5% v
This reverts commit 1e276d1a10539a8cd97d2ad63c073a9a43f0f1ef.
The poll_fd_woke() and nl_sock_woke() function added in that commit are
no longer used, so there is no reason to keep them in the tree.
---
lib/netlink-socket.c |8
lib/netlink-socket.h |1 -
lib/poll-loop.c | 33
The main goal of this series of patches is to reverse a regression in
netperf TCP_CRR performance results that was noticed in the transition
from Open vSwitch 1.2 to 1.3. The Netperf documentation describes
TCP_CRR as:
A TCP_CRR test is like a merger of a TCP_RR and TCP_CC test
wh
This optimization on its own provides about 37% benefit against a
load of a single netperf CRR test, but at the same time it penalizes
ovs-benchmark by about 11%. We can get back the CRR performance
loss, and more, other ways, so the first step is to revert this
patch.
---
lib/dpif-linux.c | 47
On Mon, Nov 21, 2011 at 4:29 AM, jamal wrote:
>
> Reading your website - it seems you indicate the code to be portable
> across other platforms. That is a noble reason - but i dont think
> it is good enough a reason in my opinion to have your own
> replicated infrastructure on top of Linux.
It's
On Mon, Nov 21, 2011 at 4:20 AM, jamal wrote:
> On Fri, 2011-11-18 at 15:12 -0800, Jesse Gross wrote:
>> Open vSwitch is a multilayer Ethernet switch targeted at virtualized
>> environments. In addition to supporting a variety of features
>> expected in a traditional hardware switch, it enables f
As they say: Only thing worse than no comment is an incorrect one.
Looks good, thanks.
Ethan
On Tue, Nov 15, 2011 at 17:17, Ben Pfaff wrote:
> ---
> vswitchd/bridge.c | 7 +++
> 1 files changed, 3 insertions(+), 4 deletions(-)
>
> diff --git a/vswitchd/bridge.c b/vswitchd/bridge.c
> ind
Looks good.
Ethan
On Tue, Nov 15, 2011 at 17:16, Ben Pfaff wrote:
> ---
> ofproto/ofproto-dpif.c | 9 +
> ofproto/ofproto-provider.h | 2 +-
> 2 files changed, 6 insertions(+), 5 deletions(-)
>
> diff --git a/ofproto/ofproto-dpif.c b/ofproto/ofproto-dpif.c
> index e89ae89..0c
I would like to see some discussion wrt. Jamal's feedback, which is that
a lot of the side-band functionality added by this code is either 1) already
doable with packet scheduler actions or 2) should be implemented there.
___
dev mailing list
dev@openvsw
On Tue, Nov 22, 2011 at 10:58 AM, Stephen Hemminger
wrote:
> On Tue, 22 Nov 2011 10:34:04 -0800
> Ben Pfaff wrote:
>
>> On Tue, Nov 22, 2011 at 10:30:05AM -0800, Jesse Gross wrote:
>> > OVS has quite a few global symbols that should be scoped with a
>> > prefix to prevent collisions with other mo
On Tue, 22 Nov 2011 10:34:04 -0800
Ben Pfaff wrote:
> On Tue, Nov 22, 2011 at 10:30:05AM -0800, Jesse Gross wrote:
> > OVS has quite a few global symbols that should be scoped with a
> > prefix to prevent collisions with other modules in the kernel.
> >
> > Suggested-by: Stephen Hemminger
> > S
On Tue, Nov 22, 2011 at 10:30:05AM -0800, Jesse Gross wrote:
> OVS has quite a few global symbols that should be scoped with a
> prefix to prevent collisions with other modules in the kernel.
>
> Suggested-by: Stephen Hemminger
> Signed-off-by: Jesse Gross
Acked-by: Ben Pfaff
I didn't apply o
On Mon, Nov 21, 2011 at 4:27 PM, Stephen Hemminger
wrote:
> One more comment...
>
> Shouldn't this device be using netdev_increment_features() like bridging and
> bonding
> to have the features of the pseudo device reflect those of the underlying
> hardware.
> This would make the device have TSO
Congratulation!!! This is to inform you that you have won £1,000,000.00 in the
Microsoft Promo End Of Year Edition, held this November 2011 in United Kingdom.
Your winning Reference # is: K4294BM Contact your Regional Claim Officer, Dr.
Maxwell Andrews with the information below.
Dr. Maxwell A
44 matches
Mail list logo