From: Joe Stringer
This patch adds a new configuration option, "max-idle" to the
Open_vSwitch "other-config" column. This sets how long datapath flows
are cached in the datapath before revalidators expire them.
Signed-off-by: Joe Stringer
Signed-off-by: Justin Pettit
---
ofproto/ofproto-dpif-
From: Joe Stringer
This should assist testing of datapath performance, as it allows us to
skip "warming up" the flow limit value.
Signed-off-by: Joe Stringer
Signed-off-by: Ben Pfaff
---
ofproto/ofproto-dpif-upcall.c | 26 ++
1 file changed, 26 insertions(+)
diff --
This patch adds a new configuration option, "max-idle" to the
Open_vSwitch "other-config" column. This sets how long datapath flows
are cached in the datapath before ovs-vswitchd thread expire them.
This commit is a backport of commit 72310b04 (upcall: Configure
datapath max-idle through ovs-vsctl
This patch adds a new configuration option, "max-idle" to the
Bridge "other-config" column. This sets how long datapath flows,
for the configured bridge, are cached in the datapath before
ovs-vswitchd thread expires them.
This commit is a backport of commit 72310b04 (upcall: Configure
datapath max
This patch adds a new configuration option, "max-idle" to the
Bridge "other-config" column. This sets how long datapath flows,
for the configured bridge, are cached in the datapath before
ovs-vswitchd thread expires them.
This commit is a backport of commit 72310b04 (upcall: Configure
datapath max
This series backports the commit 72310b04 (upcall: Configure datapath
max-idle through ovs-vsctl.) to branch-1.10 - branch-2.1, for testing
purpose.
--
1.7.9.5
___
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
ovs-vsctl settting forwarding delay does not seem to be working. Forwarding
delay is always 15 seconds.
# ovs-vsctl set bridge obr2 other_config:stp-forward-delay=0
# tcpdump -vvv -i br2
tcpdump: listening on br2, link-type EN10MB (Ethernet), capture size 65535
bytes
03:55:30.833907 STP 802.1d, Co
thanks, pushed both with the suggested fixes.
On Wed, Jun 18, 2014 at 2:46 PM, Pravin Shelar wrote:
> On Mon, Jun 16, 2014 at 1:18 PM, Andy Zhou wrote:
>> When deleting a mask from the mask array, we always move the last entry
>> into its current location. Another approach can be NULL in its
>>
This is needed to add flow dump batching to dpif-netdev, specifically
for malloc-ing 'max_flows' key and mask buffers.
Signed-off-by: Ryan Wilson
---
v2: Addressed Joe's comments, split into 2 patches
---
lib/dpif-linux.c |4 ++--
lib/dpif-netdev.c |4 ++--
lib/d
Previously, flows were retrieved one by one when dumping flows for
datapaths of type 'netdev'. This increased contention for the dump's
mutex, negatively affecting revalidator performance.
This patch retrieves batches of flows when dumping flows for datapaths
of type 'netdev'.
Signed-off-by: Ryan
> Hi Ryan, a few broad comments:
>
> Could you split this into two patches, the dpif interface change and the
> dpif-netdev implementation change?
>
>
Sure will do.
> I'm a little uneasy about the change to the dpif_flow_dump_next()
> interface, just because the size of the output parameter was p
This patch will fix the naming issue, too. But yea ideally I'd like to merge
something soon so we can have this fixed.
diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c
index fbdb6b3..0f068e9 100644
--- a/lib/netdev-dpdk.c
+++ b/lib/netdev-dpdk.c
@@ -1179,9 +1179,12 @@ dpdk_init(int argc, char *
Hi Ryan, a few broad comments:
Could you split this into two patches, the dpif interface change and the
dpif-netdev implementation change?
I'm a little uneasy about the change to the dpif_flow_dump_next()
interface, just because the size of the output parameter was previously
paired with the outp
On Thu, Jun 19, 2014 at 4:47 PM, Jesse Gross wrote:
> On Thu, Jun 19, 2014 at 4:30 PM, Pravin Shelar wrote:
>> On Thu, Jun 19, 2014 at 4:07 PM, Jesse Gross wrote:
>>> On Thu, Jun 19, 2014 at 1:33 PM, Pravin Shelar wrote:
> diff --git a/datapath/flow.c b/datapath/flow.c
> index f1bb95d..
On Thu, Jun 19, 2014 at 4:30 PM, Pravin Shelar wrote:
> On Thu, Jun 19, 2014 at 4:07 PM, Jesse Gross wrote:
>> On Thu, Jun 19, 2014 at 1:33 PM, Pravin Shelar wrote:
diff --git a/datapath/flow.c b/datapath/flow.c
index f1bb95d..7b108ed 100644
--- a/datapath/flow.c
+++ b/datapa
On Thu, Jun 19, 2014 at 4:07 PM, Jesse Gross wrote:
> On Thu, Jun 19, 2014 at 1:33 PM, Pravin Shelar wrote:
>> git am warning:
>> /home/pravin/ovs/w7/.git/rebase-apply/patch:53: trailing whitespace.
>>
>> }
>>
>> warning: 1 line adds whitespace errors.
>
> Fixed.
>
>>
On Thu, Jun 19, 2014 at 2:54 PM, Jesse Gross wrote:
> On Thu, Jun 19, 2014 at 2:01 PM, Pravin Shelar wrote:
>> On Thu, Jun 19, 2014 at 1:46 PM, Jesse Gross wrote:
>>> On Thu, Jun 19, 2014 at 1:30 PM, Pravin Shelar wrote:
On Tue, Jun 10, 2014 at 4:47 PM, Jesse Gross wrote:
> @@ -520,18
On Thu, Jun 19, 2014 at 1:33 PM, Pravin Shelar wrote:
> git am warning:
> /home/pravin/ovs/w7/.git/rebase-apply/patch:53: trailing whitespace.
>
> }
>
> warning: 1 line adds whitespace errors.
Fixed.
> ---
> compiler warning:
>
> lib/odp-ut
Large TX and RX queues are needed for high speed 10 GbE physical NICS.
Observed a 250% zero loss improvement over small NIC queue test for
A port to port flow test.
Signed-off-by: Mike A. Polehn
diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c
index fbdb6b3..d1bcc73 100644
--- a/lib/netdev-dp
Checked out the code that was modified by the patch and found that both
MAX_RX_QUEUE_LEN and
MAX_TX_QUEUE_LEN definitions are dually used for different meaning. Also the
name
implies something different then a set NIC queue size. Resubmitting patch with
zero loss gain in
comment following thi
I wrote “—dpdk argument” in the first place because other programs that make
use of rte_eal_init() do not have the “—dpdk” argument (see examples/ dir in
dpdk release).
As a matter of facts, you’re right, we’re skipping the program name. This has
some implications I didn’t think about:
- r
There is an improvement in 2544 zero loss measurements, but it takes another
patch to actually be able to get a reasonable measurement with standard test
equipment.
Should I redo it with the new enum change. I am not sure of using an enum for a
single constant.
Mike Polehn
-Original Messa
I did a bit of research and I found out what's happening here.
Here's an OVS command for DPDK:
/home/ryan/ovs/_build-gcc/vswitchd/ovs-vswitchd --dpdk -c 0x1 -n 4 --
unix:/home/alex/root/run/db.sock --pidfile --log-file --enable-dummy
-vconsole:off --detach
What will happen when running with DPDK
On Thu, Jun 19, 2014 at 2:01 PM, Pravin Shelar wrote:
> On Thu, Jun 19, 2014 at 1:46 PM, Jesse Gross wrote:
>> On Thu, Jun 19, 2014 at 1:30 PM, Pravin Shelar wrote:
>>> On Tue, Jun 10, 2014 at 4:47 PM, Jesse Gross wrote:
@@ -520,18 +534,24 @@ int ovs_flow_extract(struct sk_buff *skb, u16
Also another question. Does this patch result in a measurable
improvement in any benchmarks? If so, would you please note it in the
commit message? If not, I'm not sure we should merge this yet.
Ethan
On Thu, Jun 19, 2014 at 2:45 PM, Polehn, Mike A wrote:
> I coming from an earlier version th
I coming from an earlier version that had the arguments first setup was as a
number, then used in several
places including the tx cache size and didn't catch that new 3rd definition
were used as I moved the patch forward
to try on the latest git updates before sending.
There is also a queue s
On Mon, Jun 16, 2014 at 9:46 AM, Daniele Di Proietto
wrote:
> rte_eal_init() returns the number of parsed dpdk arguments to skip.
> dpdk_init() should add 1 to that number, because it has already skipped the
> "--dpdk" argument itself
>
> Signed-off-by: Daniele Di Proietto
Looks good.
Acked-by:
On Wed, Jun 18, 2014 at 4:54 PM, Ryan Wilson 76511 wrote:
> Well we're really not 'skipping' the '--dpdk' argument since that is
> passed to rte_eal_init() as well. We're skipping the program name which is
> the path to ovs-vswitchd. I'd change the comment in the patch to something
> like:
>
Well
On Thu, Jun 19, 2014 at 1:49 PM, Jesse Gross wrote:
> On Thu, Jun 19, 2014 at 1:30 PM, Pravin Shelar wrote:
>> On Tue, Jun 10, 2014 at 4:47 PM, Jesse Gross wrote:
>>> Some tunnel formats have mechanisms for indicating that packets are
>>> OAM frames that should be handled specially (either as hi
On Thu, Jun 19, 2014 at 1:46 PM, Jesse Gross wrote:
> On Thu, Jun 19, 2014 at 1:30 PM, Pravin Shelar wrote:
>> On Tue, Jun 10, 2014 at 4:47 PM, Jesse Gross wrote:
>>> As new protocols are added, the size of the flow key tends to
>>> increase although few protocols care about all of the fields. I
On Thu, Jun 19, 2014 at 1:31 PM, Pravin Shelar wrote:
> Git am warning.
>
> Applying: datapath: Factor out allocation and verification of actions.
>
> /home/pravin/ovs/w7/.git/rebase-apply/patch:64: trailing whitespace.
>
> warning: 1 line adds whitespace errors.
>
> On Tue, Jun 10, 2014 at 4:47 P
On Thu, Jun 19, 2014 at 1:30 PM, Pravin Shelar wrote:
> On Tue, Jun 10, 2014 at 4:47 PM, Jesse Gross wrote:
>> Some tunnel formats have mechanisms for indicating that packets are
>> OAM frames that should be handled specially (either as high priority or
>> not forwarded beyond an endpoint). This
On Thu, Jun 19, 2014 at 1:30 PM, Pravin Shelar wrote:
> On Tue, Jun 10, 2014 at 4:47 PM, Jesse Gross wrote:
>> As new protocols are added, the size of the flow key tends to
>> increase although few protocols care about all of the fields. In
>> order to optimize this for hashing and matching, OVS
The ovs-vsctl man page has some examples of configuring STP. It should be
easy to extrapolate how configure the forwarding delay with the
"stp-forward-delay" configuration field described in the
ovs-vswitchd.conf.db man page.
--Justin
On Thu, Jun 19, 2014 at 1:07 PM, Changbin Liu
wrote:
> Hi
git am warning:
/home/pravin/ovs/w7/.git/rebase-apply/patch:53: trailing whitespace.
}
warning: 1 line adds whitespace errors.
---
compiler warning:
lib/odp-util.c:869:15: warning: cast from 'uint8_t *' (aka 'unsigned
char *') to 'struct g
Git am warning.
Applying: datapath: Factor out allocation and verification of actions.
/home/pravin/ovs/w7/.git/rebase-apply/patch:64: trailing whitespace.
warning: 1 line adds whitespace errors.
On Tue, Jun 10, 2014 at 4:47 PM, Jesse Gross wrote:
> As the size of the flow key grows, it can pu
On Tue, Jun 10, 2014 at 4:47 PM, Jesse Gross wrote:
> Some tunnel formats have mechanisms for indicating that packets are
> OAM frames that should be handled specially (either as high priority or
> not forwarded beyond an endpoint). This provides support for allowing
> those types of packets to be
On Tue, Jun 10, 2014 at 4:47 PM, Jesse Gross wrote:
> Currently, the flow information that is matched for tunnels and
> the tunnel data passed around with packets is the same. However,
> as additional information is added this is not necessarily desirable,
> as in the case of pointers.
>
> This ad
On Tue, Jun 10, 2014 at 4:47 PM, Jesse Gross wrote:
> In some cases, the names of tunnel protocols are longer than the
> maximum allowed after concatenating port numbers. This truncates
> names to the correct size rather than assert-failing on the assumption
> that collisions between names of prot
On Tue, Jun 10, 2014 at 4:47 PM, Jesse Gross wrote:
> As new protocols are added, the size of the flow key tends to
> increase although few protocols care about all of the fields. In
> order to optimize this for hashing and matching, OVS uses a varible
> length portion of the key. However, when fi
Hi Folks,
I am wondering whether there is a command to set the forwarding delay of
OVS bridge. In Linux bridge, there is "brctl setfd" to set the forwarding
delay.
The motivation is that when doing live migration of a VM which is attached
to OVS bridge, I need to make sure the Gratuitous ARP will
One question: why not just increase MAX_RX_QUEUE_LEN and
MAX_TX_QUEUE_LEN instead of creating new #defines?
Just a thought. I'd like Pravin to review this as I don't know this
code as well as him.
Ethan
On Thu, Jun 19, 2014 at 9:59 AM, Polehn, Mike A wrote:
> Large TX and RX queues are needed
Large TX and RX queues are needed for high speed 10 GbE physical NICS.
Signed-off-by: Mike A. Polehn
diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c
index fbdb6b3..d1bcc73 100644
--- a/lib/netdev-dpdk.c
+++ b/lib/netdev-dpdk.c
@@ -70,6 +70,9 @@ static struct vlog_rate_limit rl = VLOG_RATE_LIM
Hey check this out!
Coupons & Cash Back Stores: All
http://www.fatwallet.com/cash-back-shopping/
Be a part of the $49,459,771 earned by cash back shoppers!
Simply shop your favorite stores through FatWallet. Earn cash back rewards on
everyday purchases. It's your cash, given back, so you always
44 matches
Mail list logo
