On Mon, Feb 5, 2024 at 1:35 PM Roberto Bartzen Acosta via dev
wrote:
>
> Updating the reference documentation with the inclusion of possible building
> problems with libjemalloc and solution suggestions.
>
> Reported-at:
> https://bugs.launchpad.net/ubuntu/+source/openvswitch/+bug/2015748
>
I mentioned Ales in my notes but forgot to CC him in my reply. Whoops.
I've added him now :)
On 2/7/24 19:22, Mark Michelson wrote:
Hi Martin,
Thanks a bunch for your first OVN contribution. I have comments below.
On 2/7/24 10:56, Martin Kalcok wrote:
When a machine from an external network
Hi Martin,
Thanks a bunch for your first OVN contribution. I have comments below.
On 2/7/24 10:56, Martin Kalcok wrote:
When a machine from an external network tries to access machine in the
OVN's internal network with SNAT enabled, using protocol like TCP, it
receives connection reset after
On 2/7/24 22:07, Adrian Moreno wrote:
>
>
> On 2/7/24 19:05, Ilya Maximets wrote:
>> On 2/5/24 13:02, Adrian Moreno wrote:
>>> netdev_linux_get_speed needs to lock netdev_linux->mutex, and so do the
>>> internal tc operations. Therefore, the former cannot be called from the
>>> latter.
>>>
>>>
On 2/7/24 19:05, Ilya Maximets wrote:
On 2/5/24 13:02, Adrian Moreno wrote:
netdev_linux_get_speed needs to lock netdev_linux->mutex, and so do the
internal tc operations. Therefore, the former cannot be called from the
latter.
Create a lock-free version of netdev_linux_get_speed() and call
On 2/5/24 13:02, Adrian Moreno wrote:
> netdev_linux_get_speed needs to lock netdev_linux->mutex, and so do the
> internal tc operations. Therefore, the former cannot be called from the
> latter.
>
> Create a lock-free version of netdev_linux_get_speed() and call it from
> tc operations.
>
>
The patch, when 'persistent' flag is specified, makes the IP selection
in a range persistent across reboots.
Signed-off-by: Paolo Valerio
---
NEWS | 3 ++-
lib/conntrack.c | 27 +--
lib/conntrack.h | 1 +
lib/dpif-netdev.c | 2 ++
4 files changed, 26
The userspace conntrack only supported hash for port selection.
With the patch, both userspace and kernel datapath support the random
flag.
The default behavior remains the same, that is, if no flags are
specified, hash is selected.
Signed-off-by: Paolo Valerio
---
Paolo Valerio writes:
> The patch, when 'persistent' flag is specified, makes the IP selection
> in a range persistent across reboots.
>
> Signed-off-by: Paolo Valerio
> ---
> NEWS | 3 ++-
> lib/conntrack.c | 26 ++
> lib/conntrack.h | 1 +
>
Considering following topology:
client - sw0 - lrp0 - lr - lrp1 - sw1 - server
sw0 in subnet 192.168.0.0/24
sw1 in subnet 172.168.0.0/24
SNAT configured for client
gateway_mtu=1400 configured for lrp0
If we send UDP traffic from client to server
and server responds with packet bigger than 1400
The ICMP need frag could be generated after routing
stage when the unSNAT already happened. Add flows
that will ensure that we are keeping track of the CT
state and do appropriate CT nat action later on.
Because the ICMP traffic is related to already existing
one in this case we can use adjusted
The ct_commit nat was hardcoded to use DNAT zone
in router pipeline. Extend it that it accepts
two new arguments (snat/dnat) which will determine
the zone for router pipeline. The switch pipeline
has only one, so it resolves to the same for both arguments.
In order to keep backward compatibility
On 06/02/2024 12:28, Ilya Maximets wrote:
> On 2/2/24 13:46, Kevin Traynor wrote:
>> Branches 2.17/3.0/3.1/3.2 are using newer DPDK LTS releases.
>>
>> Update the faq.
>>
>> Signed-off-by: Kevin Traynor
>> ---
>> Documentation/faq/releases.rst | 8
>> 1 file changed, 4 insertions(+), 4
The patch, when 'persistent' flag is specified, makes the IP selection
in a range persistent across reboots.
Signed-off-by: Paolo Valerio
---
NEWS | 3 ++-
lib/conntrack.c | 26 ++
lib/conntrack.h | 1 +
lib/dpif-netdev.c | 2 ++
4 files changed, 27
The userspace conntrack only supported hash for port selection.
With the patch, both userspace and kernel datapath support the random
flag.
The default behavior remains the same, that is, if no flags are
specified, hash is selected.
Signed-off-by: Paolo Valerio
---
On 1/24/24 15:27, Mohammad Heib wrote:
> Currently, OVN-IC doesn't support a way to tell the end-user when their
> changes
> to the IC-NB database have propagated successfully to the IC-SB Database.
>
> This patch series adds basic support for the sequence number status protocol
> that is
Bleep bloop. Greetings Martin Kalcok, 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 88 characters long (recommended limit is 79)
WARNING: Comment with 'xxx'
Bleep bloop. Greetings Ales Musil, 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.
git-am:
error: sha1 information is lacking or useless (tests/ovn.at).
error: could not build fake ancestor
Thank you Ales and Xavier.
I applied patches 1-26 to main as is. The patches to remove table
numbers from comments can be uploaded as a separate series or part of v3
for patches 27-29.
Mark Michelson
On 2/6/24 10:01, Ales Musil wrote:
On Tue, Feb 6, 2024 at 3:55 PM Xavier Simonart wrote:
When a machine from an external network tries to access machine in the
OVN's internal network with SNAT enabled, using protocol like TCP, it
receives connection reset after first packets are successfully exchanged.
This setup may be a bit unusual and requires that whatever is responsible
for
Considering following topology:
client - sw0 - lrp0 - lr - lrp1 - sw1 - server
sw0 in subnet 192.168.0.0/24
sw1 in subnet 172.168.0.0/24
SNAT configured for client
gateway_mtu=1400 configured for lrp0
If we send UDP traffic from client to server
and server responds with packet bigger than 1400
The ct_commit nat was hardcoded to use DNAT zone
in router pipeline. Extend it that it accepts
two new arguments (snat/dnat) which will determine
the zone for router pipeline. The switch pipeline
has only one, so it resolves to the same for both arguments.
In order to keep backward compatibility
The ICMP need frag could be generated after routing
stage when the unSNAT already happened. Add flows
that will ensure that we are keeping track of the CT
state and do appropriate CT nat action later on.
Because the ICMP traffic is related to already existing
one in this case we can use adjusted
The macro had two issues:
- It never used the third (optional) argument (command such as trim_zeros).
- The default command was wrong, causing the test to always succeed.
Fixes: 9857ef8f61a0 ("tests: fixed multiple tests not properly waiting for
packets to be received")
Signed-off-by: Xavier
Open vSwitch module accepts actions as a list from the netlink socket
and then creates a copy which it uses in the action set processing.
During processing of the action list on a packet, the module keeps a
count of the execution depth and exits processing if the action depth
goes too high.
Add a test case into the netlink checks that will show the number of
nested action recursions won't exceed 16. Going to 17 on a small
clone call isn't enough to exhaust the stack on (most) systems, so
it should be safe to run even on systems that don't have the fix
applied.
Signed-off-by: Aaron
The ovs module allows for some actions to recursively contain an action
list for complex scenarios, such as sampling, checking lengths, etc.
When these actions are copied into the internal flow table, they are
evaluated to validate that such actions make sense, and these calls
happen recursively.
On 5 Feb 2024, at 13:36, Roberto Bartzen Acosta via dev wrote:
> Updating the reference documentation with the inclusion of possible building
> problems with libjemalloc and solution suggestions.
>
> Reported-at:
> https://bugs.launchpad.net/ubuntu/+source/openvswitch/+bug/2015748
>
On Tue, Feb 06, 2024 at 11:15:24AM -0500, Mike Pattrick wrote:
> Include inner offsets in functions where l3 and l4 offsets are either
> modified or checked.
>
> Fixes: 084c8087292c ("userspace: Support VXLAN and GENEVE TSO.")
> Signed-off-by: Mike Pattrick
...
> diff --git
On Mon, Feb 05, 2024 at 09:36:14AM -0300, Roberto Bartzen Acosta via dev wrote:
> Updating the reference documentation with the inclusion of possible building
> problems with libjemalloc and solution suggestions.
>
> Reported-at:
>
On Mon, Feb 05, 2024 at 01:02:10PM +0100, Adrian Moreno wrote:
> netdev_linux_get_speed needs to lock netdev_linux->mutex, and so do the
> internal tc operations. Therefore, the former cannot be called from the
> latter.
>
> Create a lock-free version of netdev_linux_get_speed() and call it from
31 matches
Mail list logo