With the current OVS offload design, when an offload-device fails to add a
flow rule and returns an error, OVS adds the rule to the kernel datapath.
The flow gets processed by the kernel datapath for the entire life of that
flow. This is fine when an error is returned by the device due to lack of
s
This is the first patch in the patch-set to support dynamic rebalancing
of offloaded flows.
The patch detects OOR condition on a netdev port when ENOSPC error is
returned by TC-Flower while adding a flow rule. A new structure is added
to the netdev called "netdev_hw_info", to store OOR related inf
This is the second patch in the patch-set to support dynamic rebalancing
of offloaded flows.
The packets-per-second (pps) rate for each flow is computed in the context
of revalidator threads when the flow stats are retrieved. The pps-rate is
computed only after a flow is revalidated and is not sch
This is the third patch in the patch-set to support dynamic rebalancing
of offloaded flows.
The dynamic rebalancing functionality is implemented in this patch. The
ukeys that are not scheduled for deletion are obtained and passed as input
to the rebalancing routine. The rebalancing is done in the
This is the fourth patch in the patch-set to support dynamic rebalancing
of offloaded flows.
A new OVS configuration parameter "offload-rebalance", is added to ovsdb.
The default value of this is "disable". To enable this feature, set the
value of this parameter to "pps-rate", which provides packe
Hey Aaron,
On Wed, 11 Jul 2018 15:55:34 -0400
0-day Robot wrote:
> Bleep bloop. Greetings Jakub Sitnicki, I am a robot and I have tried out
> your patch.
> Thanks for your contribution.
>
> I encountered some error that I wasn't expecting. See the details below.
>
>
> build:
> depbase=`ech
On Wed, 11 Jul 2018 16:04:32 -0400
0-day Robot wrote:
> Bleep bloop. Greetings Jakub Sitnicki, I am a robot and I have tried out
> your patch.
> Thanks for your contribution.
>
> I encountered some error that I wasn't expecting. See the details below.
>
>
> checkpatch:
> WARNING: Line is 82
RE: RECOVERED FUND.
In reference to the signal received from our Investigative Team concerning
overdue payments. We write to inform you that some funds have been recovered
from some Security Companies, and sub financial houses, meant to be paid out to
owners but diverted.
However, it was fo
Jakub Sitnicki writes:
> Hey Aaron,
>
> On Wed, 11 Jul 2018 15:55:34 -0400
> 0-day Robot wrote:
>
>> Bleep bloop. Greetings Jakub Sitnicki, I am a robot and I have tried out
>> your patch.
>> Thanks for your contribution.
>>
>> I encountered some error that I wasn't expecting. See the detail
0-day Robot writes:
> Bleep bloop. Greetings William Tu, I am a robot and I have tried out
> your patch.
> Thanks for your contribution.
>
> I encountered some error that I wasn't expecting. See the details below.
>
>
> checkpatch:
> ERROR: Too many signoffs; are you missing Co-authored-by line
Jakub Sitnicki writes:
> On Wed, 11 Jul 2018 16:04:32 -0400
> 0-day Robot wrote:
>
>> Bleep bloop. Greetings Jakub Sitnicki, I am a robot and I have
>> tried out your patch.
>> Thanks for your contribution.
>>
>> I encountered some error that I wasn't expecting. See the details below.
>>
>>
On 7/11/2018 7:23 PM, Tiago Lam wrote:
From: Mark Kavanagh
There are numerous factors that must be considered when calculating
the size of an mbuf:
- the data portion of the mbuf must be sized in accordance With Rx
buffer alignment (typically 1024B). So, for example, in order to
successfu
This series extends ovn-nbctl tool with support for the daemon mode, where
ovn-nbctl acts a long-lived process that accepts commands over a UNIX socket.
The daemon can be started the same way as any other OVS/OVN server:
ovn-nbctl --detach --pidfile --log-file
While commands can be issued to it
Having a constant in addition to the constant expression for the default
table style allows us to reset 'struct table_style' variables to default
style.
Signed-off-by: Jakub Sitnicki
---
lib/table.h | 2 ++
1 file changed, 2 insertions(+)
diff --git a/lib/table.h b/lib/table.h
index 313ac1dd2..
If IDL was created with monitoring and alerts turned on by default for
all columns, then there is no harm in allowing the API users to ask
again for monitoring and alerts to be enabled for any given column.
This allows us to run prerequisites handlers for db-ctl and ovn-nbctl
commands once the IDL
Split out a routine for the main ovn-nbctl loop.
Preparatory work for introducing daemon mode.
Signed-off-by: Jakub Sitnicki
---
ovn/utilities/ovn-nbctl.c | 20
1 file changed, 16 insertions(+), 4 deletions(-)
diff --git a/ovn/utilities/ovn-nbctl.c b/ovn/utilities/ovn-nbct
Destroy commands in the same routine where they were allocated.
Preparatory work for reusing the main loop in daemon mode.
Signed-off-by: Jakub Sitnicki
---
ovn/utilities/ovn-nbctl.c | 13 +++--
1 file changed, 7 insertions(+), 6 deletions(-)
diff --git a/ovn/utilities/ovn-nbctl.c b/ov
Destroy IDL resources in the routine where we allocated them.
Preparatory work for reusing the main loop in daemon mode.
Signed-off-by: Jakub Sitnicki
---
ovn/utilities/ovn-nbctl.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/ovn/utilities/ovn-nbctl.c b/ovn/utilities/o
Introduce an output parameter for the flag that signals need to retry
running the command. This leaves the return value for error reporting.
Preparatory work for reusing the main loop in daemon mode.
Signed-off-by: Jakub Sitnicki
---
ovn/utilities/ovn-nbctl.c | 20 +---
1 file c
Instead of terminating the process, return the error to the caller.
This will allow us to reuse the main loop in daemon mode.
Signed-off-by: Jakub Sitnicki
---
ovn/utilities/ovn-nbctl.c | 46 --
1 file changed, 32 insertions(+), 14 deletions(-)
diff
Let the caller handle the errors instead of reporting it and
terminating. Prepare for reusing the main loop in daemon mode.
Signed-off-by: Jakub Sitnicki
---
ovn/utilities/ovn-nbctl.c | 19 +--
1 file changed, 13 insertions(+), 6 deletions(-)
diff --git a/ovn/utilities/ovn-nbctl
Instead of terminating the process, return the error to the caller.
This will allow us to reuse the prerequisites runner in daemon mode.
Signed-off-by: Jakub Sitnicki
---
ovn/utilities/ovn-nbctl.c | 21 +++--
1 file changed, 15 insertions(+), 6 deletions(-)
diff --git a/ovn/uti
Extend the main loop and the command runner so that the caller can
specify a timeout for poll_block(). This will allow us to break out of
the main loop when waiting on IDL, like in the blocked '--wait=sb/hv
sync' case.
Signed-off-by: Jakub Sitnicki
---
ovn/utilities/ovn-nbctl.c | 27
This will allow us to direct oneline-formatted output to other sinks
than stdout if needed. Preparatory work for daemon mode.
Signed-off-by: Jakub Sitnicki
---
ovn/utilities/ovn-nbctl.c | 53 ++-
1 file changed, 34 insertions(+), 19 deletions(-)
diff
Make ovn-nbctl act as a unixctl server if we were asked to detach. This
turns ovn-nbctl into a long-lived process that acts a proxy for
interacting with NB DB. The main difference to regular mode of ovn-nbctl
is that in the daemon mode, a local copy of database contents has to be
obtained only once
Signed-off-by: Jakub Sitnicki
---
tests/ovn-nbctl.at | 21 +
1 file changed, 21 insertions(+)
diff --git a/tests/ovn-nbctl.at b/tests/ovn-nbctl.at
index 73a61a4be..89daf631a 100644
--- a/tests/ovn-nbctl.at
+++ b/tests/ovn-nbctl.at
@@ -1382,3 +1382,24 @@ inactivity_probe:
Signed-off-by: Jakub Sitnicki
---
tests/ovn.at | 19 +++
1 file changed, 19 insertions(+)
diff --git a/tests/ovn.at b/tests/ovn.at
index d1a8967dd..adb99db77 100644
--- a/tests/ovn.at
+++ b/tests/ovn.at
@@ -10541,3 +10541,22 @@ OVN_CHECK_PACKETS([hv2/vif2-tx.pcap], [vif2.expected
Signed-off-by: Jakub Sitnicki
---
tests/ovn-nbctl.at | 21 +
1 file changed, 21 insertions(+)
diff --git a/tests/ovn-nbctl.at b/tests/ovn-nbctl.at
index 89daf631a..60b4d0c9c 100644
--- a/tests/ovn-nbctl.at
+++ b/tests/ovn-nbctl.at
@@ -1403,3 +1403,24 @@ AT_CHECK([ovn-nbctl ls
Currently, rule_insert() API doesnot have return value. There are some possible
scenarios where rule insertions can fail at run-time even though the static
checks during rule_construct() had passed previously.
Some possible scenarios for failure of rule insertions:
**) Rule insertions can fail dyn
> Currently, rule_insert() API doesnot have return value. There are some
possible
> scenarios where rule insertions can fail at run-time even though the
static
> checks during rule_construct() had passed previously.
> Some possible scenarios for failure of rule insertions:
> **) Rule insertions ca
Aravind Prasad S writes:
> Currently, rule_insert() API doesnot have return value. There are some
> possible
/doesnot/doesn't/ or /doesnot/does not/
Or maybe even just remove that sentence.
:)
> scenarios where rule insertions can fail at run-time even though the static
> checks during rul
On 12/07/2018 14:37, Ian Stokes wrote:
> On 7/11/2018 7:23 PM, Tiago Lam wrote:
>> From: Mark Kavanagh
>>
>> There are numerous factors that must be considered when calculating
>> the size of an mbuf:
>> - the data portion of the mbuf must be sized in accordance With Rx
>>buffer alignment (typ
On 7/12/2018 4:40 PM, Lam, Tiago wrote:
On 12/07/2018 14:37, Ian Stokes wrote:
On 7/11/2018 7:23 PM, Tiago Lam wrote:
From: Mark Kavanagh
There are numerous factors that must be considered when calculating
the size of an mbuf:
- the data portion of the mbuf must be sized in accordance With Rx
On 12/07/2018 17:07, Ian Stokes wrote:
> On 7/12/2018 4:40 PM, Lam, Tiago wrote:
>> On 12/07/2018 14:37, Ian Stokes wrote:
>>> On 7/11/2018 7:23 PM, Tiago Lam wrote:
From: Mark Kavanagh
There are numerous factors that must be considered when calculating
the size of an mbuf:
>>>
Hello Ian,
Document was very helpful. My testbed has OVS-DPDK vhost and virito in a
container, it had enough information for me to get it working inside the
container.
I have one question for you, I run 'iperf3' tcp and udp performance tests
for more than 5 minutes and packet traverses through vh
This patch only affects rhel6 spec file.
The rhel6 kmod spec file is renamed from openvswitch-kmod-rhel6.spec
to kmod-openvswitch-rhel6.spec . This is to prepare for the next
patches to support building multiple kernel versions in the main
package. The rename makes the spec file consistent with th
This patch only affects rhel6 spec file.
Previouly the kernel_module_package macro is used to generate spec file
template to build kmod-openvswitch RPM. The main package only contains
the openvswitch.conf for depmod. The macro is now removed. Everything is
built in the main package instead. This e
This patch only affects rhel6 spec file.
RHEL 7.4 introduced backward incompatible changes in the kernel. As
a result, prebuilt PRM packages against kernels newer than 693.17.1
will cannot be used on systems with older kernels, vice versa.
This patch allows multiple kernel version numbers delimit
Currently, rule_insert() API does not have return value. There are some possible
scenarios where rule insertions can fail at run-time even though the static
checks during rule_construct() had passed previously.
Some possible scenarios for failure of rule insertions:
**) Rule insertions can fail dyn
Hi Aaron,
Thanks a lot for the review.
Submitted the patch with comments addressed as below.
> /doesnot/doesn't/ or /doesnot/does not/
> Or maybe even just remove that sentence.
> :)
Addressed. Changed to "does not"
> This second if looks strange. I think it should be removed.
Addressed.
>
On Thu, Jul 12, 2018 at 10:35:54AM -0700, Martin Xu wrote:
> This patch only affects rhel6 spec file.
>
> The rhel6 kmod spec file is renamed from openvswitch-kmod-rhel6.spec
> to kmod-openvswitch-rhel6.spec . This is to prepare for the next
> patches to support building multiple kernel versions i
A common issue is users pairing the incorrect version of OVS to DPDK
when working outside of the build tree.
To avoid this this commit updates the OVS DPDK documentation to explicitly
flag that users should consult the OVS to DPDK release mapping in FAQ if
working outside of the OVS build tree.
S
On Thu, Jul 12, 2018 at 10:35:55AM -0700, Martin Xu wrote:
> This patch only affects rhel6 spec file.
>
> Previouly the kernel_module_package macro is used to generate spec file
> template to build kmod-openvswitch RPM. The main package only contains
> the openvswitch.conf for depmod. The macro is
On Thu, Jul 12, 2018 at 10:35:56AM -0700, Martin Xu wrote:
> This patch only affects rhel6 spec file.
>
> RHEL 7.4 introduced backward incompatible changes in the kernel. As
> a result, prebuilt PRM packages against kernels newer than 693.17.1
> will cannot be used on systems with older kernels, v
On 07/12/2018 09:40 AM, Jakub Sitnicki wrote:
If IDL was created with monitoring and alerts turned on by default for
all columns, then there is no harm in allowing the API users to ask
again for monitoring and alerts to be enabled for any given column.
This allows us to run prerequisites handler
On 07/12/2018 09:40 AM, Jakub Sitnicki wrote:
Make ovn-nbctl act as a unixctl server if we were asked to detach. This
turns ovn-nbctl into a long-lived process that acts a proxy for
interacting with NB DB. The main difference to regular mode of ovn-nbctl
is that in the daemon mode, a local copy o
This is tough. The problem with a test like this is that you can't
really tell that the sync is working as intended. It's possible that the
ovn-nbctl is not actually waiting, but by coincidence, the database has
its values updated before you check them.
You'd need something internal to ovn-nbc
Thanks, applied to master.
On Wed, Jul 11, 2018 at 01:25:03PM -0700, Yifeng Sun wrote:
> Thanks.
>
> Reviewed-by: Yifeng Sun
>
> On Mon, Jul 9, 2018 at 2:37 PM, Ben Pfaff wrote:
>
> > OVS trips over this warning all over the place, so it's not worth leaving
> > on.
> >
> > Signed-off-by: Ben
On Thu, Jul 12, 2018 at 07:21:36AM -0400, Aaron Conole wrote:
> Aaron Conole writes:
>
> > Ben Pfaff writes:
> >
> >> The bot couldn't apply this patch presumably because it tried to apply
> >> it against master. Maybe the bot should notice version numbers in [] in
> >> subject lines and try ag
On Wed, Jul 11, 2018 at 10:33:51AM -0700, Ravi Kerur wrote:
> On Tue, Jul 10, 2018 at 2:07 PM Ben Pfaff wrote:
>
> > On Thu, Jul 05, 2018 at 01:44:33PM -0700, Ravi Kerur wrote:
> > > During scale flow add/delete (>10K), I am seeing ovs-vswitchd cpu usage
> > > spike to 100% and stay there without
When I run this test with DPDK enabled, it fails because it ends up using
a different recirculation ID when DPDK is not enabled. I guess that's a
little weird but the recirculation IDs are not supposed to be significant,
so this change makes the test ignore it.
Signed-off-by: Ben Pfaff
---
test
uint8_t values shouldn't be passed to ntohs().
Found by soarse.
Signed-off-by: Ben Pfaff
---
lib/netdev-dpdk.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c
index bb4d60f26cc4..b4ed4ad5919c 100644
--- a/lib/netdev-dpdk.c
+++ b/lib/
Neither of these is a real problem.
Signed-off-by: Ben Pfaff
---
lib/netdev-dpdk.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c
index b4ed4ad5919c..d485a53dacf1 100644
--- a/lib/netdev-dpdk.c
+++ b/lib/netdev-dpdk.c
@@ -4384,7 +4
Signed-off-by: Ben Pfaff
---
lib/netdev-dpdk.c | 16
1 file changed, 4 insertions(+), 12 deletions(-)
diff --git a/lib/netdev-dpdk.c b/lib/netdev-dpdk.c
index d485a53dacf1..9bf21856075b 100644
--- a/lib/netdev-dpdk.c
+++ b/lib/netdev-dpdk.c
@@ -4079,12 +4079,8 @@ dump_flow_patte
The macros are hard to read. This makes it a little more readable.
Signed-off-by: Ben Pfaff
---
configure.ac | 1 +
lib/netdev-dpdk.c | 235 --
lib/netdev-dummy.c| 134
lib/netdev-linux.c| 340 +++---
This allows netdev-dpdk.c to compile cleanly with sparse.
Signed-off-by: Ben Pfaff
---
Makefile.am |2 +-
build-aux/initial-tab-whitelist |1 +
include/sparse/automake.mk |9 +
include/sparse/rte_byteorder.h | 281
include/sparse/rte_esp.h|
Thanks. It worked like a charm.
Tested-by: Yifeng Sun
Reviewed-by: Yifeng Sun
On Tue, Jul 10, 2018 at 1:40 PM, Ben Pfaff wrote:
> This was useful for testing commit 4fe080160685 ("flow: Fix buffer overread
> for crafted IPv6 packets.").
>
> Signed-off-by: Ben Pfaff
> ---
> v1->v2: Updated c
On Thu, Jul 12, 2018 at 12:23 PM, Flavio Leitner wrote:
> On Thu, Jul 12, 2018 at 10:35:55AM -0700, Martin Xu wrote:
> > This patch only affects rhel6 spec file.
> >
> > Previouly the kernel_module_package macro is used to generate spec file
> > template to build kmod-openvswitch RPM. The main pa
This patch only affects rhel6 spec file.
The rhel6 kmod spec file is renamed from openvswitch-kmod-rhel6.spec
to kmod-openvswitch-rhel6.spec . This is to prepare for the next
patches to support building multiple kernel versions in the main
package. The rename makes the spec file consistent with th
This patch only affects rhel6 spec file.
RHEL 7.4 introduced backward incompatible changes in the kernel. As
a result, prebuilt PRM packages against kernels newer than 693.17.1
will cannot be used on systems with older kernels, vice versa.
This patch allows multiple kernel version numbers delimit
This patch only affects rhel6 spec file.
Previouly the kernel_module_package macro is used to generate spec file
template to build kmod-openvswitch RPM. The main package only contains
the openvswitch.conf for depmod. The macro is now removed. Everything is
built in the main package instead. This e
Thanks, applied to master.
On Thu, Jul 12, 2018 at 03:11:14PM -0700, Yifeng Sun wrote:
> Thanks. It worked like a charm.
>
> Tested-by: Yifeng Sun
>
> Reviewed-by: Yifeng Sun
>
> On Tue, Jul 10, 2018 at 1:40 PM, Ben Pfaff wrote:
>
> > This was useful for testing commit 4fe080160685 ("flow:
This still needs a review.
On Thu, Jun 14, 2018 at 02:49:23PM -0700, Ben Pfaff wrote:
> These don't trigger any new actual warnings in my own build.
>
> GCC 8.x adds other new warning options that are enabled by -Wall or
> -Wextra. This commit doesn't explicitly enable those because OVS already
On Fri, Jun 15, 2018 at 04:29:22PM -0700, Ben Pfaff wrote:
> ofpacts_check__() was a huge switch statement with special cases for many
> different kinds of actions. This made it unwieldy and put the special
> cases far away from the rest of the code related to a given action. This
> commit refact
This series still needs reviews.
On Fri, Jun 15, 2018 at 03:11:09PM -0700, Ben Pfaff wrote:
> When ovsdb-server closes a remote connection, it logs a message about it
> that includes the reason. Until now this has included sessions that it
> closes when it exits. That meant that, when --run was
On Thu, Jul 12, 2018 at 07:48:42PM +0100, Ian Stokes wrote:
> A common issue is users pairing the incorrect version of OVS to DPDK
> when working outside of the build tree.
>
> To avoid this this commit updates the OVS DPDK documentation to explicitly
> flag that users should consult the OVS to DP
On Thu, Jul 12, 2018 at 03:17:27PM -0700, Martin Xu wrote:
> On Thu, Jul 12, 2018 at 12:23 PM, Flavio Leitner wrote:
>
> > On Thu, Jul 12, 2018 at 10:35:55AM -0700, Martin Xu wrote:
> > > This patch only affects rhel6 spec file.
[...]
> > > +
> > > +# By default, build against the latest installe
On Wed, Jul 11, 2018 at 09:05:11AM -0700, Darrell Ball wrote:
> When conn_update_state() returns true, conn has been freed, so skip calling
> handle_ftp_ctl() with this conn and instead follow code path for new
> connections.
>
> Fixes: bd5e81a0e596 ("Userspace Datapath: Add ALG infra and FTP.")
>
On Thu, Jul 12, 2018 at 04:25:22PM -0700, Martin Xu wrote:
> This patch only affects rhel6 spec file.
>
> The rhel6 kmod spec file is renamed from openvswitch-kmod-rhel6.spec
> to kmod-openvswitch-rhel6.spec . This is to prepare for the next
> patches to support building multiple kernel versions i
On Thu, Jul 12, 2018 at 08:54:26PM -0300, Flavio Leitner wrote:
> On Thu, Jul 12, 2018 at 04:25:22PM -0700, Martin Xu wrote:
> > This patch only affects rhel6 spec file.
> >
> > The rhel6 kmod spec file is renamed from openvswitch-kmod-rhel6.spec
> > to kmod-openvswitch-rhel6.spec . This is to pre
On Wed, Jul 11, 2018 at 2:49 PM, Yifeng Sun wrote:
> Hi Darrell,
>
> Thanks for the test. It shouldn't be that slow. For the same
> 'nsh -forward' test, it is pretty fast on my test machine.
>
> ## -- ##
> ## openvswitch 2.9.90 test suite. ##
> ## -
Hi Aaron/Ben/All,
If possible, Kindly review the patch and let me know
your views.
On Thu, Jul 12, 2018 at 11:34 PM, Aravind Prasad S
wrote:
> Currently, rule_insert() API does not have return value. There are some
> possible
> scenarios where rule insertions can fail at run-time even though th
Hi Darrell,
Thanks. There are really lots of things to consider. I will come up with v2.
Best,
Yifeng
On Thu, Jul 12, 2018 at 7:14 PM, Darrell Ball wrote:
>
>
> On Wed, Jul 11, 2018 at 2:49 PM, Yifeng Sun
> wrote:
>
>> Hi Darrell,
>>
>> Thanks for the test. It shouldn't be that slow. For the
73 matches
Mail list logo