Re: [ovs-dev] [PATCH ovn v7 2/3] Support LSP:options:requested-chassis as a list

Bleep bloop. Greetings Ihar Hrachyshka, 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=`echo utilities/ovn-appctl.o | sed 's|[^/]*$|.deps/&|;s|\.o$||'`;\ gcc -std=gnu99

[ovs-dev] [PATCH ovn v7 3/3] Clone packets to both port chassis

When multiple chassis are set in requested-chassis, port binding is configured in multiple cluster locations. In case of live migration scenario, only one of the locations run a workload at a particular point in time. Yet, it's expected that the workload may switch to running at an additional

[ovs-dev] [PATCH ovn v7 2/3] Support LSP:options:requested-chassis as a list

When the option is set to a comma separated list of chassis names, OVN will attempt to bind the port at any number of other locations in addition to the main chassis. This is useful in live migration scenarios where it's important to prepare the environment for workloads to move to, avoiding

[ovs-dev] [PATCH ovn v7 1/3] Update port-up on main chassis only

In a future patch, there will be a scenario where the same port has attachments at multiple (specifically, 2) chassis, so make sure that 'up' property is updated by the main chassis only. Signed-off-by: Ihar Hrachyshka --- controller/binding.c| 9 ++--- controller/binding.h

[ovs-dev] [PATCH ovn v7 0/3] Support multiple requested-chassis

This version excludes the implementation for rarp activation strategy. This last piece needs some more work, sending an updated series without it to speed up reviews and merges. This iteration of the patch series includes the following changes: v7: dropped the patch that tags all traffic from

Re: [ovs-dev] [PATCH ovn v6 3/5] Support LSP:options:requested-chassis as a list

Thanks Han for your comments. I've handled what you spotted + added more scenarios, for 3 chassis, switching roles of chassis from main to additional and back, to have some chassis not claiming the binding. I hope this resolves your concerns. Ihar On Wed, May 11, 2022 at 2:22 AM Han Zhou

Re: [ovs-dev] [PATCH ovn v6 1/5] Tag all packets that arrived from a tunnel as LOCAL_ONLY

This patch won't be needed after the next iteration of 'Clone packets...' patch in the set drops. I'm dropping it in the next set. On Wed, May 11, 2022 at 2:20 AM Han Zhou wrote: > > > > On Thu, May 5, 2022 at 6:38 AM Ihar Hrachyshka wrote: > > > > A next patch may need to clone packets to

Re: [ovs-dev] [PATCH ovn v6 2/5] Update port-up on main chassis only

Correct, I'm handling this in the next iteration. On Wed, May 11, 2022 at 2:23 AM Han Zhou wrote: > > > > On Thu, May 5, 2022 at 6:38 AM Ihar Hrachyshka wrote: > > > > In a future patch, there will be a scenario where the same port has > > attachments at multiple (specifically, 2) chassis, so

Re: [ovs-dev] [PATCH ovn v6 4/5] Clone packets to both port chassis

Thanks for spotting unneeded flows. These are artifacts from swapping the order of local/remote handling. (I've also noticed that now we don't need to tag packets coming from tunnels as LOCAL_ONLY to make it work, so I'll also drop the patch from the set.) As for convoluted logic in

Re: [ovs-dev] [PATCH ovn] Allow for setting the Next server IP in the DHCP header

On Wed, May 11, 2022 at 11:34 AM wrote: > > From: Lucas Alvares Gomes > > In order to PXE boot a baremetal server using the OVN DHCP server we > need to allow users to set the "next-server" (siaddr) [0] field in the > DHCP header. > > While investigating this issue by comparing the DHCPOFFER and

Re: [ovs-dev] [PATCH ovn v2] ovn-controller: Consider zone 0 as a valid zone when restoring.

On Wed, May 18, 2022 at 4:41 PM Mark Michelson wrote: > > Looks good to me Dumitru. > > Acked-by: Mark Michelson Thanks Dumiru and Mark. I applied this patch to the main branch and backported to branch-22.03 and branch-21.12. Numan > > On 5/18/22 13:47, Dumitru Ceara wrote: > > 0 is a valid

Re: [ovs-dev] [PATCH ovn v2] ovn-controller: Consider zone 0 as a valid zone when restoring.

Looks good to me Dumitru. Acked-by: Mark Michelson On 5/18/22 13:47, Dumitru Ceara wrote: 0 is a valid zone ID and some CMSs might actually use it. Reported-at: https://bugzilla.redhat.com/show_bug.cgi?id=2087194 Signed-off-by: Dumitru Ceara --- v2: Use str_to_uint() as suggested by Ilya.

Re: [ovs-dev] [PATCH ovn v3] ovn-controller: Set Port_Binding.up field only if the Southbound DB is aware of this field

Thank you for the update, Mary. I have pushed the change to main, branch-22.03, and branch-21.12. On 5/18/22 14:21, mary.mano...@nutanix.com wrote: From: Mary Manohar This patch is fixing a backward compatibility issue when using an older ovn-northd (20.09). The 20.09 Southbound schema is

Re: [ovs-dev] [PATCH ovn 0/3] Adding generic port security flows.

On 5/18/22 14:17, Numan Siddique wrote: On Mon, May 16, 2022 at 5:42 PM Numan Siddique > wrote: On Mon, May 16, 2022 at 4:27 PM Mark Michelson mailto:mmich...@redhat.com>> wrote: Hi Numan, I've taken a close look at the patches in this series,

Re: [ovs-dev] [PATCH ovn] northd: Honor ct-snat-zone option for common case.

On 5/17/22 16:48, Numan Siddique wrote: On Fri, May 6, 2022 at 9:47 AM Mark Michelson > wrote: Commit 4deac4509 introduced the new "czone" variants for ct_dnat and ct_snat. This makes it so that a single conntrack zone is used for both DNAT and SNAT on

Re: [ovs-dev] [PATCH v2] sha1: Use implementation from openssl if available.

On 5/18/22 20:00, Ilya Maximets wrote: > On 5/16/22 09:40, Dumitru Ceara wrote: >> On 5/7/22 00:55, Ilya Maximets wrote: >>> Implementation of SHA1 in OpenSSL library is much faster and optimized >>> for all available CPU architectures and instruction sets. OVS should >>> use it instead of

Re: [ovs-dev] [PATCH ovn v2] Do not set 'up' attribute in the port-binding table if ovn-northd is at an older version.

Hi Numan Sure, I have sent out a v3 with the updated commit message. Thanks Mary From: Numan Siddique Date: Wednesday, May 18, 2022 at 7:37 AM To: Mark Michelson Cc: Mary Manohar , Xavier Simonart , ovs-dev@openvswitch.org Subject: Re: [ovs-dev] [PATCH ovn v2] Do not set 'up' attribute in

[ovs-dev] [PATCH ovn v3] ovn-controller: Set Port_Binding.up field only if the Southbound DB is aware of this field

From: Mary Manohar This patch is fixing a backward compatibility issue when using an older ovn-northd (20.09). The 20.09 Southbound schema is not aware of the 'up' field in the Port_Binding table. ovn-controller should set Port_Binding.up field only if the Southbound DB is aware of this field

Re: [ovs-dev] [PATCH ovn 0/3] Adding generic port security flows.

On Mon, May 16, 2022 at 5:42 PM Numan Siddique wrote: > > > On Mon, May 16, 2022 at 4:27 PM Mark Michelson > wrote: > >> Hi Numan, >> >> I've taken a close look at the patches in this series, and they seem >> really good. They're very well commented and well tested as well. It's >> quite easy

Re: [ovs-dev] [PATCH v2] sha1: Use implementation from openssl if available.

On 5/16/22 09:40, Dumitru Ceara wrote: > On 5/7/22 00:55, Ilya Maximets wrote: >> Implementation of SHA1 in OpenSSL library is much faster and optimized >> for all available CPU architectures and instruction sets. OVS should >> use it instead of internal implementation if possible. >> >>

Re: [ovs-dev] [PATCH ovn] nb: Add Load_Balancer.options:neighbor_responder knob.

I applied this to main. On 5/11/22 14:48, Mark Michelson wrote: Thanks for this Dumitru. Acked-by: Mark Michelson On 4/28/22 11:33, Dumitru Ceara wrote: This allows CMS to tweak the way logical routers reply to ARP/ND packets targeting load balancer VIPs.  By default a router only replies

Re: [ovs-dev] [PATCH ovn] ovn-controller: Consider zone 0 as a valid zone when restoring.

On 5/18/22 18:41, Ilya Maximets wrote: > On 5/18/22 18:09, Dumitru Ceara wrote: >> 0 is a valid zone ID and some CMSs might actually use it. >> >> Reported-at: https://bugzilla.redhat.com/show_bug.cgi?id=2087194 >> Signed-off-by: Dumitru Ceara >> --- >> controller/ovn-controller.c | 15

[ovs-dev] [PATCH ovn v2] ovn-controller: Consider zone 0 as a valid zone when restoring.

0 is a valid zone ID and some CMSs might actually use it. Reported-at: https://bugzilla.redhat.com/show_bug.cgi?id=2087194 Signed-off-by: Dumitru Ceara --- v2: Use str_to_uint() as suggested by Ilya. --- controller/ovn-controller.c | 15 ++- 1 file changed, 10 insertions(+), 5

Re: [ovs-dev] [PATCH ovn] ovn-controller: Consider zone 0 as a valid zone when restoring.

On 5/18/22 18:09, Dumitru Ceara wrote: > 0 is a valid zone ID and some CMSs might actually use it. > > Reported-at: https://bugzilla.redhat.com/show_bug.cgi?id=2087194 > Signed-off-by: Dumitru Ceara > --- > controller/ovn-controller.c | 15 ++- > 1 file changed, 10 insertions(+), 5

Re: [ovs-dev] OVN - flushing conntrack entries for newly allocated zones

On 5/17/22 16:10, Dumitru Ceara wrote: > Hi Justin, > > I know it's a while ago since this patch [0][1] made it in OVN but I was > wondering if you remembered some of the reasons behind doing this. > > More recently, OVN got a new feature to allow the CMS to explicitly > specify a zone ID to be

[ovs-dev] [PATCH ovn] ovn-controller: Consider zone 0 as a valid zone when restoring.

0 is a valid zone ID and some CMSs might actually use it. Reported-at: https://bugzilla.redhat.com/show_bug.cgi?id=2087194 Signed-off-by: Dumitru Ceara --- controller/ovn-controller.c | 15 ++- 1 file changed, 10 insertions(+), 5 deletions(-) diff --git

Re: [ovs-dev] [PATCH V2 3/3] netdev-offload-dpdk: Add geneve header pattern match

1. The focus was VXLAN. Need to enhance DPDK to support it. If this is not important, let’s abandon this patch-set until DPDK is enhanced. 2. There is no need. DPDK has only specific encaps for VXLAN/NVGRE. Other encaps are done only by “RAW”. In OVS VXLAN is used if applicable, and

Re: [ovs-dev] [PATCH V2 3/3] netdev-offload-dpdk: Add geneve header pattern match

Eli, I'm trying to understand options handling during Geneve encap/decap offload. 1. This patchset will allow decap offload for Geneve w/o options only. I do not think that covers important use cases for Geneve. Geneve was meant to be used with options. Why is the limitation of not

Re: [ovs-dev] [PATCH ovn v2] Do not set 'up' attribute in the port-binding table if ovn-northd is at an older version.

On Tue, May 17, 2022 at 9:55 AM Mark Michelson wrote: > OK thanks for the explanations, Xavier and Mary. This is pretty > important since we expect that ovn-controller should be upgraded before > ovn-northd. > > I'm acking the patch because the content is correct. I suggest that when > this is

Re: [ovs-dev] [PATCH ovn] utilities: nbctl: dump lbs in load_balancer group running {ls, lr}-lb-list

On Tue, May 10, 2022 at 4:22 PM Mark Michelson wrote: > Thanks for this, Lorenzo. > > Acked-by: Mark Michelson > Thanks. I applied this patch to the main branch. Numan > > On 4/27/22 06:47, Lorenzo Bianconi wrote: > > Dump load balancers belonging to the load_balancer groups attached to the

[ovs-dev] [PATCH v4 1/2] dpif-netdev : Fix ALB parameters type mismatch.

From: Lin Huang The ALB parameters should never be negative. So it's to use smap_get_ulonglong() or smap_get_uint() to get it properly. Fixes: 5bf84282482a ("Adding support for PMD auto load balancing") Signed-off-by: Lin Huang --- lib/dpif-netdev.c | 15 --- 1 file changed, 8

[ovs-dev] [PATCH v4 2/2] dpif-netdev : Fix ALB 'rebalance_intvl' max hard limit.

From: Lin Huang Currently the pmd-auto-lb-rebal-interval's value was not been checked properly. It maybe a negative, or too big value (>2 weeks between rebalances), which will be lead to a big unsigned value. So reset it to default if the value exceeds the max permitted as described in

[ovs-dev] [PATCH v4 0/2] Fix ALB parameters type and value mismatch.

From: Lin Huang This series fix ALB parameters type mismatch and set a hard limit to rebalance_intvl. It also includes some self-tests. Changes since v3: 1. Fix CI warnings Changes since v2: 1. Fix type mismatch int dpif-netdev.c. Changes since v1: 1. Fix the commit message and

Re: [ovs-dev] [PATCH v3 0/2] Fix ALB parameters type and value mismatch.

Hi Kevin, Thanks for the review. It's very helpful to me. > -Original Message- > From: Kevin Traynor > Sent: Tuesday, May 17, 2022 11:34 PM > To: mit...@outlook.com; ovs-dev@openvswitch.org > Cc: Lin Huang > Subject: Re: [PATCH v3 0/2] Fix ALB parameters type and value mismatch. > >

Re: [ovs-dev] [PATCH] dpif-netdev: Only hash port number when necessary.

> -Original Message- > From: Ilya Maximets > Sent: Wednesday 18 May 2022 11:35 > To: Mike Pattrick ; Ferriter, Cian > Cc: ovs-dev@openvswitch.org; i.maxim...@ovn.org > Subject: Re: [ovs-dev] [PATCH] dpif-netdev: Only hash port number when > necessary. > > On 5/10/22 17:03, Mike

Re: [ovs-dev] [PATCH v6 0/7] Fix UndefinedBehaviorSanitizer reported issues and enable it in CI.

On 5/18/22 12:31, Ilya Maximets wrote: > On 4/11/22 13:37, Dumitru Ceara wrote: >> As privately reported by Aaron Conole, and by Jeffrey Walton [0] >> there's currently a number of undefined behavior instances in >> the OVS code base. Running the OVS (and OVN) tests with UBSan [1] >> enabled

Re: [ovs-dev] [PATCH ovn v4] northd: add configuration option for enabling parallelization

On 5/18/22 14:59, Dumitru Ceara wrote: > On 5/18/22 12:07, Xavier Simonart wrote: >> This patch is intended to change the way to enable northd lflow build >> parallelization, as well as enable runtime change of number of threads. >> Before this patch, the following was needed to use

Re: [ovs-dev] [PATCH ovn v4] northd: add configuration option for enabling parallelization

On 5/18/22 12:07, Xavier Simonart wrote: > This patch is intended to change the way to enable northd lflow build > parallelization, as well as enable runtime change of number of threads. > Before this patch, the following was needed to use parallelization: > - enable parallelization through

Re: [ovs-dev] [PATCH v1] Pmd.at: fix dpcls and dpif configuration test cases.

> -Original Message- > From: Amber, Kumar > Sent: Tuesday 17 May 2022 18:35 > To: ovs-dev@openvswitch.org > Cc: i.maxim...@ovn.org; Ferriter, Cian ; Stokes, Ian > ; Phelan, Michael ; Van > Haaren, Harry ; Amber, Kumar > > Subject: [PATCH v1] Pmd.at: fix dpcls and dpif configuration

Re: [ovs-dev] [PATCH] dynamic-string: Fix undefined behavior due to offsetting null pointer.

On 5/17/22 22:52, Ilya Maximets wrote: > On 4/26/22 17:16, Dumitru Ceara wrote: >> On 4/26/22 16:29, Adrian Moreno wrote: >>> >>> >>> On 4/12/22 15:29, Dumitru Ceara wrote: When compiled with '-fsanitize=undefined' running 'ovsdb-client --timestamp monitor Open_vSwitch' in a sandbox

[ovs-dev] [PATCH v2] dynamic-string: Fix undefined behavior due to offsetting null pointer.

When compiled with clang and '-fsanitize=undefined' set, running 'ovsdb-client --timestamp monitor Open_vSwitch' in a sandbox triggers the following undefined behavior (flagged by UBSan): lib/dynamic-string.c:207:38: runtime error: applying zero offset to null pointer #0 0x4ebc18 in

Re: [ovs-dev] [PATCH v2] tests: Properly kill ovsdb test processes

On 5/10/22 20:33, Rosemarie O'Riorden wrote: > The FreeBSD CI builds keep failing because processes of tests are not > properly killed. This leaves the build hanging until it times out, and > ultimately fails. > > Changes name of pidfile pid2 to 2.pid so that the > on_exit 'kill `cat *.pid`' will

Re: [ovs-dev] [PATCH] dpif-netdev: Only hash port number when necessary.

On 5/10/22 17:03, Mike Pattrick wrote: > On Fri, Apr 29, 2022 at 11:44 AM Cian Ferriter > wrote: >> >> The hash of the port number is only needed when a DPCLS needs to be >> created. Move the hash calculation inside the if to accomplish this. >> >> Signed-off-by: Cian Ferriter >> --- > > Good

Re: [ovs-dev] [PATCH v2] ovs-save: Get highest ofp version error.

On 5/11/22 10:53, Han Ding wrote: > When setting just one ofp version to protocols of bridge, The function > get_highest_ofp_version in ovs-save parse it error. > > For example: > $ ovs-vsctl get bridge br-int protocols > [OpenFlow15] > > $ ovs-vsctl get bridge br-int protocols | > sed

Re: [ovs-dev] [PATCH v6 0/7] Fix UndefinedBehaviorSanitizer reported issues and enable it in CI.

On 4/11/22 13:37, Dumitru Ceara wrote: > As privately reported by Aaron Conole, and by Jeffrey Walton [0] > there's currently a number of undefined behavior instances in > the OVS code base. Running the OVS (and OVN) tests with UBSan [1] > enabled uncovers these. > > This series fixes the issues

Re: [ovs-dev] [PATCH ovn v4] northd: add configuration option for enabling parallelization

Bleep bloop. Greetings Xavier Simonart, 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 lacks whitespace around operator WARNING: Line lacks whitespace around

[ovs-dev] [PATCH ovn v4] northd: add configuration option for enabling parallelization

This patch is intended to change the way to enable northd lflow build parallelization, as well as enable runtime change of number of threads. Before this patch, the following was needed to use parallelization: - enable parallelization through use_parallel_build in NBDB - use --dummy-numa to select

Re: [ovs-dev] [PATCH v2] tests/mfex: Improve pcap script for mfex tests.

Hi Eelco, Please find my replies Inline. > -Original Message- > From: Eelco Chaudron > Sent: Thursday, May 12, 2022 2:53 PM > To: Amber, Kumar > Cc: ovs-dev@openvswitch.org; Ferriter, Cian ; > Stokes, Ian ; Van Haaren, Harry > ; Ilya Maximets > Subject: Re: [PATCH v2] tests/mfex:

[ovs-dev] [PATCH v3] tests/mfex: Improve pcap script for mfex tests.

The mfex pcap generation script is improved for varied length traffic and also removes the hard coded mfex_pcap and instead uses the script itself to generate complex traffic patterns for testing. Signed-off-by: Kumar Amber Acked-by: Cian Ferriter --- v3: - Fix comments(Eelco). - Script

Re: [ovs-dev] [PATCH v5 5/5] acinclude: Add seperate checks for AVX512 ISA.

> -Original Message- > From: Pai G, Sunil > Sent: Tuesday 17 May 2022 15:52 > To: Ferriter, Cian ; ovs-dev@openvswitch.org > Cc: Van Haaren, Harry > Subject: RE: [PATCH v5 5/5] acinclude: Add seperate checks for AVX512 ISA. > > > Checking for each of the required AVX512 ISA

Re: [ovs-dev] [v4 7/8] netdev-offload-tc: Offloading rules with police actions

On 18 May 2022, at 9:46, Jianbo Liu wrote: > On Wed, 2022-05-18 at 09:19 +0200, Eelco Chaudron wrote: >> >> >> On 18 May 2022, at 8:54, Jianbo Liu wrote: >> >>> On Wed, 2022-05-18 at 08:25 +0200, Eelco Chaudron wrote: On 18 May 2022, at 5:27, Jianbo Liu wrote: > On Mon,

Re: [ovs-dev] [v4 7/8] netdev-offload-tc: Offloading rules with police actions

On Wed, 2022-05-18 at 09:19 +0200, Eelco Chaudron wrote: > > > On 18 May 2022, at 8:54, Jianbo Liu wrote: > > > On Wed, 2022-05-18 at 08:25 +0200, Eelco Chaudron wrote: > > > > > > > > > On 18 May 2022, at 5:27, Jianbo Liu wrote: > > > > > > > On Mon, 2022-05-16 at 13:34 +0200, Eelco

Re: [ovs-dev] [v4 7/8] netdev-offload-tc: Offloading rules with police actions

On 18 May 2022, at 8:54, Jianbo Liu wrote: > On Wed, 2022-05-18 at 08:25 +0200, Eelco Chaudron wrote: >> >> >> On 18 May 2022, at 5:27, Jianbo Liu wrote: >> >>> On Mon, 2022-05-16 at 13:34 +0200, Eelco Chaudron wrote: On 3 May 2022, at 5:08, Jianbo Liu via dev wrote: >

Re: [ovs-dev] [v4 7/8] netdev-offload-tc: Offloading rules with police actions

On Wed, 2022-05-18 at 08:25 +0200, Eelco Chaudron wrote: > > > On 18 May 2022, at 5:27, Jianbo Liu wrote: > > > On Mon, 2022-05-16 at 13:34 +0200, Eelco Chaudron wrote: > > > > > > > > > On 3 May 2022, at 5:08, Jianbo Liu via dev wrote: > > > > > > > When offloading rule, tc should be filled

Re: [ovs-dev] [v4 7/8] netdev-offload-tc: Offloading rules with police actions

On 18 May 2022, at 5:27, Jianbo Liu wrote: > On Mon, 2022-05-16 at 13:34 +0200, Eelco Chaudron wrote: >> >> >> On 3 May 2022, at 5:08, Jianbo Liu via dev wrote: >> >>> When offloading rule, tc should be filled with police index, instead >>> of meter id. As meter is mapped to police action, and