On Mar 02, Simon Horman wrote:
> On Thu, Mar 02, 2023 at 01:03:32PM +0100, Lorenzo Bianconi wrote:
> > Drop ip packets with ct status set to invalid in post snat and
> > lb_aff_learn router stages.
> > Skip ICMPv{4,6} error messages packet in ct.inv rules in order to avoid
> > to introduce too
On 3/2/23 05:24, Dumitru Ceara wrote:
On 3/2/23 11:15, Ilya Maximets wrote:
On 3/1/23 23:36, Ilya Maximets wrote:
It's common to have 'hairpin_snat_ip' to be configured for each and
every Load Balancer in a setup. For example, ovn-kubernetes does that,
setting '169.254.169.5 fd69::5' value
On Thu, Mar 02, 2023 at 03:57:42PM +0800, Faicker Mo wrote:
> The device may be deleted and added with ifindex changed.
> The tc rules on the device will be deleted if the device is deleted.
> The func tc_del_filter will fail when flow del. The mapping of
> ufid to tc will not be deleted.
> The
On Thu, Mar 02, 2023 at 01:03:32PM +0100, Lorenzo Bianconi wrote:
> Drop ip packets with ct status set to invalid in post snat and
> lb_aff_learn router stages.
> Skip ICMPv{4,6} error messages packet in ct.inv rules in order to avoid
> to introduce too complicated code.
>
> Reported-at:
On Thu, Mar 02, 2023 at 09:08:49AM +0100, Ales Musil wrote:
> There is a bug in userspace datapath that causes
> failures within the LB affinity test [0].
> Skip the test until the bug is fixed.
>
> [0] https://bugzilla.redhat.com/2170828
> Tested-by: Simon Horman
> Signed-off-by: Ales Musil
On Thu, Mar 02, 2023 at 09:08:51AM +0100, Ales Musil wrote:
> The test with jemalloc and clang is redundant as
> we are already testing jemalloc with gcc. Replace
> it with system-test-userspace which runs system tests
> over userspace OvS datapath.
>
> Tested-by: Simon Horman
> Signed-off-by:
On Thu, Mar 02, 2023 at 09:08:47AM +0100, Ales Musil wrote:
> The ADD_INT does not work very well with userspace datapath.
> When we move the interface into separate namespace
> the ovs-vswitch cannot get ethtool info from that interface
> and reports errors for some calls:
>
>
On Thu, Mar 02, 2023 at 09:08:46AM +0100, Ales Musil wrote:
> Newer versions of ping6 print some info to
> stderr when specified with -v flag. This info
> causes tests to fail. Do not use the verbose
> output.
>
> It reports following info into stderr
> which is then represtend as error by the
>
On 3/2/23 09:45, Simon Horman wrote:
On Thu, Mar 02, 2023 at 02:14:24PM +0100, Dumitru Ceara wrote:
On 3/2/23 14:11, Ilya Maximets wrote:
This patch set optimizes usage of string operations and avoids parsing
same LSP addresses multiple times.
Performance tests with ovn-heater show 5-10%
On Thu, Mar 02, 2023 at 02:14:24PM +0100, Dumitru Ceara wrote:
> On 3/2/23 14:11, Ilya Maximets wrote:
> > This patch set optimizes usage of string operations and avoids parsing
> > same LSP addresses multiple times.
> >
> > Performance tests with ovn-heater show 5-10% improvement in high-scale
>
On Thu, Mar 02, 2023 at 02:11:09PM +0100, Ilya Maximets wrote:
> For non-dynamic address, current version performs 1 strcmp + 4 attempts
> for ovs_scan. Updated code performs 1 strncmp + 1 ovs_scan.
>
> Also, for ports with both IPv4 and IPv6 specified, new version doesn't
> re-parse IPv4 twice.
On 2/28/23 11:01, Simon Horman wrote:
> On Mon, Feb 27, 2023 at 08:54:23PM +0100, Ilya Maximets wrote:
>> The only user of this function was removed in a cited commit. And
>> it is unlikely to be used in the future, since we have feature flags
>> in the database instead.
>>
>> Fixes: 3013c2869696
On 2/24/23 12:01, Felix Hüttner via dev wrote:
> the comments did refer to tables 1 below the actual table number.
>
> Signed-off-by: Felix Huettner
> ---
> northd/northd.c | 34 +-
> 1 file changed, 17 insertions(+), 17 deletions(-)
>
> diff --git
On 3/1/23 09:18, Simon Horman wrote:
> On Tue, Feb 21, 2023 at 01:38:56PM +0100, Ales Musil wrote:
>> The tests are looking for specific output from tcpdump
>> counting number of occurrences and expecting only 1. It
>> might happen that in the time period there are already two
>> or more packets
On 2/22/23 15:24, Simon Horman wrote:
> On Wed, Feb 22, 2023 at 01:56:51PM +0100, Ales Musil wrote:
>> Nothing is being freed wherever we are calling
>> ctl_fatal which is fine because the program is
>> about to shutdown anyway however one of the
>> leaks was caught by address sanitizer.
>> Fix
On 2/21/23 13:26, Ales Musil wrote:
> On Fri, Feb 17, 2023 at 9:41 PM Dumitru Ceara wrote:
>
>> As mentioned in the structure definition in lb.h, northd should never
>> use 'vip_port' instead it should always use 'port_str'. That's because
>> it might not always be possible to evaluate the port
Bleep bloop. Greetings Ilya Maximets, 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: Comment with 'xxx' marker
#52 FILE: lib/ovn-util.c:123:
/* Either
On 3/2/23 14:11, Ilya Maximets wrote:
> This patch set optimizes usage of string operations and avoids parsing
> same LSP addresses multiple times.
>
> Performance tests with ovn-heater show 5-10% improvement in high-scale
> density-light scenarios.
>
> Version 2:
> - Reduced code duplication
For non-dynamic address, current version performs 1 strcmp + 4 attempts
for ovs_scan. Updated code performs 1 strncmp + 1 ovs_scan.
Also, for ports with both IPv4 and IPv6 specified, new version doesn't
re-parse IPv4 twice.
Signed-off-by: Ilya Maximets
---
lib/ovn-util.c | 40
At the point of IPAM configuration all the addresses are already parsed.
No need to waste time parsing them again.
Reviewed-by: Simon Horman
Acked-by: Mark Michelson
Signed-off-by: Ilya Maximets
---
northd/northd.c | 54 -
1 file changed, 17
This patch set optimizes usage of string operations and avoids parsing
same LSP addresses multiple times.
Performance tests with ovn-heater show 5-10% improvement in high-scale
density-light scenarios.
Version 2:
- Reduced code duplication in is_dynamic_lsp_address(). [Mark]
Ilya Maximets
In many cases OVN components are using strlen() to check if the string
is not empty. In that case, unnecessary scan of the whole string is
performed and the length is calculated, while it's enough to just check
the first byte.
Some functions also have completely unnecessary strlen() calls where
Bleep bloop. Greetings Lorenzo Bianconi, 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
Drop ip packets with ct status set to invalid in post snat and
lb_aff_learn router stages.
Skip ICMPv{4,6} error messages packet in ct.inv rules in order to avoid
to introduce too complicated code.
Reported-at: https://bugzilla.redhat.com/show_bug.cgi?id=2160685
Signed-off-by: Lorenzo Bianconi
On Thu, Mar 2, 2023 at 12:31 PM Ales Musil wrote:
> Currently, there is one ct.est flow per LB VIP,
> that was required to keep track if we need to
> pass the "skip_snat" or "force_snat" flags.
> However since c1d6b8ac ("northd: Store skip_snat and force_snat in
> ct_label/mark")
> the flags are
Currently, there is one ct.est flow per LB VIP,
that was required to keep track if we need to
pass the "skip_snat" or "force_snat" flags.
However since c1d6b8ac ("northd: Store skip_snat and force_snat in
ct_label/mark")
the flags are carried in the ct entry and
we can use match on them the same
On 3/2/23 11:15, Ilya Maximets wrote:
> On 3/1/23 23:36, Ilya Maximets wrote:
>> It's common to have 'hairpin_snat_ip' to be configured for each and
>> every Load Balancer in a setup. For example, ovn-kubernetes does that,
>> setting '169.254.169.5 fd69::5' value unconditionally for every LB:
>>
On 3/1/23 23:36, Ilya Maximets wrote:
> It's common to have 'hairpin_snat_ip' to be configured for each and
> every Load Balancer in a setup. For example, ovn-kubernetes does that,
> setting '169.254.169.5 fd69::5' value unconditionally for every LB:
>
>
On Thu, Mar 02, 2023 at 08:51:18AM +0100, Simon Horman wrote:
> On Wed, Mar 01, 2023 at 02:27:29PM -0800, Han Zhou wrote:
> > On Wed, Mar 1, 2023 at 1:50 AM Simon Horman
> > wrote:
> > >
> > > On Wed, Feb 22, 2023 at 10:35:25PM -0800, Han Zhou wrote:
> > >
> > > Please add a patch description
On 3/2/23 01:54, Numan Siddique wrote:
> On Wed, Mar 1, 2023 at 7:25 PM Dumitru Ceara wrote:
>>
>> This is to match the ovn-kubernetes commit
>> https://github.com/ovn-org/ovn-kubernetes/commit/b5b61bc64aae
>>
>> Without it building the ovn-kubernetes container image fails because the
>> libovsdb
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.
checkpatch:
WARNING: Line is 81 characters long (recommended limit is 79)
#28 FILE:
The test with jemalloc and clang is redundant as
we are already testing jemalloc with gcc. Replace
it with system-test-userspace which runs system tests
over userspace OvS datapath.
Tested-by: Simon Horman
Signed-off-by: Ales Musil
---
v3: rebase on top of current main.
---
To be consistent moe the LB affinity tests into system-ovn-kmod
for now as they are waiting currently not working with
userspace datapath.
Signed-off-by: Ales Musil
---
tests/system-ovn-kmod.at | 595 +++
tests/system-ovn.at | 595
The dpctl/del-flows shouldn't be used on running
ofproto layer, use revalidator/purge instead.
Reviewed-by: Simon Horman
Tested-by: Simon Horman
Signed-off-by: Ales Musil
---
v3: Rebase on top of current main.
---
tests/system-ovn.at | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff
In order to allow there are ome fixes done to the
tests itself. The LB affinity needs a fix from OvS
side so it is marked as skip for the time being.
To save some on the CI the last patch replaces
unit tests with Clang and Jemalloc with the
userspace system tests, the Jemalloc still
run with GCC.
There is a bug in userspace datapath that causes
failures within the LB affinity test [0].
Skip the test until the bug is fixed.
[0] https://bugzilla.redhat.com/2170828
Tested-by: Simon Horman
Signed-off-by: Ales Musil
---
v3: Rebase on top of current main.
---
tests/system-kmod-macros.at
Add support for running system tests with
OvS userspace netdev.
Reviewed-by: Simon Horman
Signed-off-by: Ales Musil
---
v3: Rebase on top of current main.
---
.ci/linux-build.sh | 69 ++
1 file changed, 46 insertions(+), 23 deletions(-)
diff --git
The ADD_INT does not work very well with userspace datapath.
When we move the interface into separate namespace
the ovs-vswitch cannot get ethtool info from that interface
and reports errors for some calls:
|00089|netdev_linux|WARN|ethtool command ETHTOOL_GSET on network device vm1
failed: No
Bump ovs submodule to b72a7f92 which is after
ed44eefb ("conntrack: Properly unNAT inner header of related traffic.").
This patch is needed for "LB - ICMP related traffic" system test to work
with OvS userspace netdev.
Reviewed-by: Simon Horman
Tested-by: Simon Horman
Signed-off-by: Ales Musil
Newer versions of ping6 print some info to
stderr when specified with -v flag. This info
causes tests to fail. Do not use the verbose
output.
It reports following info into stderr
which is then represtend as error by the
NS_CHECK_EXEC:
ping6: sock4.fd: -1 (socktype: 0), sock6.fd: 3 (socktype:
40 matches
Mail list logo
