Re: [ovs-dev] [PATCH] ofproto-dpif-upcall: Improve concurrency by adjust flow-limit

At 2021-05-12 04:00:30, "Ben Pfaff" wrote: >On Tue, May 11, 2021 at 03:52:00PM +0800, Tao YunXiang wrote: >> Author: Tao YunXiang > >You don't need an Author: tag because Git stores the author in a >separate field. > >> +/* Use duration as a reference, adjust the number of

Re: [ovs-dev] [RFC 0/3] dpif-netlink: Introduce per-cpu upcall dispatching

Hi, On Fri, Apr 30, 2021 at 11:31:26AM -0400, Mark Gray wrote: > This series proposes a new method of distributing upcalls > to user space threads attempting to resolve a number of > issues with the current method. > I ran some tests with old V10, current master and this RFC including the

[ovs-dev] [PATCH ovn v1] Fix Segmentation Fault when LR requests duplicate requested-tnl-key

From: Karthik Chandrashekar Fixes: b6c349a271363faf277041b230f2244489c7e7a3 Signed-off-by: Karthik Chandrashekar --- northd/ovn-northd.c | 4 ++-- tests/ovn-northd.at | 18 ++ 2 files changed, 20 insertions(+), 2 deletions(-) diff --git a/northd/ovn-northd.c

[ovs-dev] bounce for Tao YunXiang

It seems that email that I send to Tao YunXiang bounces; at least, this one did. I hope that he sees my previous email via the list. - Forwarded message from Mail Delivery System - Date: Tue, 11 May 2021 20:00:42 + (UTC) From: Mail Delivery System To: b...@ovn.org Subject:

Re: [ovs-dev] [PATCH] ofproto-dpif-upcall: Improve concurrency by adjust flow-limit

On Tue, May 11, 2021 at 03:52:00PM +0800, Tao YunXiang wrote: > Author: Tao YunXiang You don't need an Author: tag because Git stores the author in a separate field. > +/* Use duration as a reference, adjust the number of flow_limit, > + * when the duration is small,

Re: [ovs-dev] [PATCH ovn v4] ovn-controller: Fix port group conjunction flow explosion problem.

On 5/6/21 6:03 PM, Han Zhou wrote: On Thu, May 6, 2021 at 1:18 PM Mark Michelson > wrote: > > Thanks, Han! > > Acked-by: Mark Michelson > Thanks Mark. I applied it to master branch. I'd also like to backport it to branch-21.03. Do

Re: [ovs-dev] [PATCH ovn v2] if-status: Add OVS interface status management module.

Hi Dumitru, The change looks good to me. It's very easy to understand the state machine, and it's easy to follow how the code is structured. Yay! I would have liked to see some more test cases, but given the discussion you had on v1 of this patch, I can understand why they're not present in

Re: [ovs-dev] [PATCH V2 1/1] dpdk: Add debug appctl to get malloc statistics.

On 11/05/2021 14:22, Eli Britstein wrote: > > On 5/11/2021 4:13 PM, Eelco Chaudron wrote: >> External email: Use caution opening links or attachments >> >> >> On 11 May 2021, at 14:53, Eli Britstein wrote: >> >>> On 5/11/2021 2:29 PM, Eelco Chaudron wrote: External email: Use caution opening

Re: [ovs-dev] [PATCH] github: Fix up malformed /etc/hosts.

On 5/11/21 4:52 PM, Ilya Maximets wrote: > On 5/11/21 4:41 PM, Aaron Conole wrote: >> Ilya Maximets writes: >> >>> For some reason /etc/hosts in GHA now contains a plain text line >>> like this: >>> >>> Note: Don't Delete this file. Also, don't remove this line. ... >>> >>> This breaks

Re: [ovs-dev] [PATCH] doc: automake: Add support for sphinx 4.0.

On 5/10/21 5:57 PM, Ilya Maximets wrote: > File layout for man pages in sphinx 4 by default changed [1] from: > > Documentation/_ref/man/page.section > > to: > > Documentation/_ref/man/section/page.section > > Ajusting our build scripts so they will be able to locate files > in new

Re: [ovs-dev] [PATCH] cirrus: Look up existing versions of python dependencies.

On 5/6/21 2:27 PM, Ilya Maximets wrote: > FreeBSD sometimes changes the base version of python3 that is > used for packages. This affects package names. For example, > currently CI is broken, because there is no more py37- versions > of sphinx and openssl available, only py38- ones: > > pkg:

Re: [ovs-dev] [PATCH ovn] github: Fix up malformed /etc/hosts.

On Tue, May 11, 2021 at 10:34 AM Dumitru Ceara wrote: > > On 5/11/21 4:23 PM, Ilya Maximets wrote: > > For some reason /etc/hosts in GHA now contains a plain text line > > like this: > > > > Note: Don't Delete this file. Also, don't remove this line. ... > > > > This breaks libunbound and makes

Re: [ovs-dev] [PATCH] github: Fix up malformed /etc/hosts.

On 5/11/21 4:41 PM, Aaron Conole wrote: > Ilya Maximets writes: > >> For some reason /etc/hosts in GHA now contains a plain text line >> like this: >> >> Note: Don't Delete this file. Also, don't remove this line. ... >> >> This breaks libunbound and makes a series of unit tests to emit >>

Re: [ovs-dev] [PATCH v2] conntrack: add coverage counters for L3 bad checksum

Paolo Valerio writes: > similarly to what already exists for L4, add conntrack_l3csum_err > and ipf_l3csum_err for L3. > > Received packets with L3 bad checksum will increase respectively > ipf_l3csum_err if they are fragments and conntrack_l3csum_err > otherwise. > > Although the patch

Re: [ovs-dev] [PATCH] github: Fix up malformed /etc/hosts.

Ilya Maximets writes: > For some reason /etc/hosts in GHA now contains a plain text line > like this: > > Note: Don't Delete this file. Also, don't remove this line. ... > > This breaks libunbound and makes a series of unit tests to emit > following warning: > |1|dns_resolve|WARN|Failed

Re: [ovs-dev] [PATCH ovn] github: Fix up malformed /etc/hosts.

On 5/11/21 4:23 PM, Ilya Maximets wrote: > For some reason /etc/hosts in GHA now contains a plain text line > like this: > > Note: Don't Delete this file. Also, don't remove this line. ... > > This breaks libunbound and makes a series of unit tests to emit > following warning: >

[ovs-dev] [PATCH ovn] github: Fix up malformed /etc/hosts.

For some reason /etc/hosts in GHA now contains a plain text line like this: Note: Don't Delete this file. Also, don't remove this line. ... This breaks libunbound and makes a series of unit tests to emit following warning: |1|dns_resolve|WARN|Failed to read etc/hosts: syntax error

Re: [ovs-dev] [PATCH] doc: automake: Add support for sphinx 4.0.

Ilya Maximets writes: > File layout for man pages in sphinx 4 by default changed [1] from: > > Documentation/_ref/man/page.section > > to: > > Documentation/_ref/man/section/page.section > > Ajusting our build scripts so they will be able to locate files > in new places. This fixes our CI

Re: [ovs-dev] [PATCH] github: Fix up malformed /etc/hosts.

On Tue, May 11, 2021 at 3:21 PM Ilya Maximets wrote: > > For some reason /etc/hosts in GHA now contains a plain text line > like this: > > Note: Don't Delete this file. Also, don't remove this line. ... > > This breaks libunbound and makes a series of unit tests to emit > following warning: >

Re: [ovs-dev] [PATCH V2 1/1] dpdk: Add debug appctl to get malloc statistics.

On 11 May 2021, at 15:22, Eli Britstein wrote: > On 5/11/2021 4:13 PM, Eelco Chaudron wrote: >> External email: Use caution opening links or attachments >> >> >> On 11 May 2021, at 14:53, Eli Britstein wrote: >> >>> On 5/11/2021 2:29 PM, Eelco Chaudron wrote: External email: Use caution

Re: [ovs-dev] [PATCH V2 1/1] dpdk: Add debug appctl to get malloc statistics.

On 5/11/2021 4:13 PM, Eelco Chaudron wrote: External email: Use caution opening links or attachments On 11 May 2021, at 14:53, Eli Britstein wrote: On 5/11/2021 2:29 PM, Eelco Chaudron wrote: External email: Use caution opening links or attachments On 2 May 2021, at 9:13, Eli Britstein

[ovs-dev] [PATCH] github: Fix up malformed /etc/hosts.

For some reason /etc/hosts in GHA now contains a plain text line like this: Note: Don't Delete this file. Also, don't remove this line. ... This breaks libunbound and makes a series of unit tests to emit following warning: |1|dns_resolve|WARN|Failed to read etc/hosts: syntax error

Re: [ovs-dev] [PATCH V2 1/1] dpdk: Add debug appctl to get malloc statistics.

On 11 May 2021, at 14:53, Eli Britstein wrote: > On 5/11/2021 2:29 PM, Eelco Chaudron wrote: >> External email: Use caution opening links or attachments >> >> >> On 2 May 2021, at 9:13, Eli Britstein wrote: >> >>> New appctl 'dpdk/get-malloc-stats' implemented to get result of >>>

Re: [ovs-dev] [PATCH V2 1/1] dpdk: Add debug appctl to get malloc statistics.

On 5/11/2021 2:29 PM, Eelco Chaudron wrote: External email: Use caution opening links or attachments On 2 May 2021, at 9:13, Eli Britstein wrote: New appctl 'dpdk/get-malloc-stats' implemented to get result of 'rte_malloc_dump_stats()' function. Could be used for debugging. This patch

Re: [ovs-dev] [PATCH] Fix redundant datapath set ethernet action with NSH Decap

On 27 Apr 2021, at 14:42, Martin Varghese wrote: > From: Martin Varghese > > When a decap action is applied on NSH header encapsulatiing a > ethernet packet a redundant set mac address action is programmed > to the datapath. Patch looks good to me, small style nit below. Acked-by: Eelco

Re: [ovs-dev] [PATCH v2] conntrack: add coverage counters for L3 bad checksum

On 30 Apr 2021, at 19:12, Paolo Valerio wrote: > similarly to what already exists for L4, add conntrack_l3csum_err > and ipf_l3csum_err for L3. > > Received packets with L3 bad checksum will increase respectively > ipf_l3csum_err if they are fragments and conntrack_l3csum_err > otherwise. > >

Re: [ovs-dev] [PATCH ovn v3] ovn-northd.c: Don't use DPG for MC_UNKNOWN related lflows.

On Tue, May 11, 2021 at 4:41 AM Dumitru Ceara wrote: > > On 5/10/21 11:59 PM, Han Zhou wrote: > > In commit b468c2c1 it avoided using DPG for multicast related lflows, but > > commit dd94f126 update the MC_UNKNOWN related lflows with DPG used again > > by mistake. This patch fixes it to avoid

Re: [ovs-dev] [PATCH ovn v2] northd: Support flow offloading for logical switches with no ACLs.

On Mon, May 10, 2021 at 11:43 AM Dumitru Ceara wrote: > > On 5/10/21 5:39 PM, Dumitru Ceara wrote: > > On 5/7/21 4:42 PM, num...@ovn.org wrote: > >> From: Numan Siddique > >> > >> Some smart NICs can't offload datapath flows matching on conntrack > >> fields. If a deployment desires to make use

Re: [ovs-dev] [v2 v2 0/6] MFEX Infrastructure + Optimizations

> -Original Message- > From: Timothy Redaelli > Sent: Monday, May 10, 2021 6:43 PM > To: Amber, Kumar ; d...@openvswitch.org > Cc: i.maxim...@ovn.org; jhs...@redhat.com; f...@redhat.com; Van Haaren, Harry > > Subject: Re: [ovs-dev] [v2 v2 0/6] MFEX Infrastructure + Optimizations > > >

Re: [ovs-dev] [PATCH V2 1/1] dpdk: Add debug appctl to get malloc statistics.

On 2 May 2021, at 9:13, Eli Britstein wrote: > New appctl 'dpdk/get-malloc-stats' implemented to get result of > 'rte_malloc_dump_stats()' function. > > Could be used for debugging. This patch looks good, however, my suggestion on the first patchset was to include the

Re: [ovs-dev] [PATCH] doc: automake: Add support for sphinx 4.0.

On Mon, May 10, 2021 at 5:57 PM Ilya Maximets wrote: > > File layout for man pages in sphinx 4 by default changed [1] from: > > Documentation/_ref/man/page.section > > to: > > Documentation/_ref/man/section/page.section > > Ajusting our build scripts so they will be able to locate files > in

Re: [ovs-dev] [PATCH] ofproto-dpif-upcall: Improve concurrency by adjust flow-limit

Bleep bloop. Greetings Tao YunXiang, 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 85 characters long (recommended limit is 79) #46 FILE:

Re: [ovs-dev] [PATCH ovn v3] ovn-northd.c: Don't use DPG for MC_UNKNOWN related lflows.

On 5/10/21 11:59 PM, Han Zhou wrote: > In commit b468c2c1 it avoided using DPG for multicast related lflows, but > commit dd94f126 update the MC_UNKNOWN related lflows with DPG used again > by mistake. This patch fixes it to avoid problems when a lflow using > MC_UNKNOWN is monitored by a

[ovs-dev] [PATCH] ofproto-dpif-upcall: Fix race condition while purging

From: Jianbo Liu There is a race condidtion between purger and handler. Handler may create new ukey and install it while executing 'ovs-appctl revalidator/purge' command. However, before handler calls transition_ukey() in handle_upcalls(), purger may get this ukey from umap, then evict and

Re: [ovs-dev] [PATCH v12 00/11] Add offload support for sFlow

On 4/27/2021 9:23 AM, Chris Mi wrote: On 3/24/2021 8:14 PM, Ilya Maximets wrote: On 3/24/21 10:17 AM, Chris Mi wrote: On 3/23/2021 10:24 PM, Ilya Maximets wrote: On 3/5/21 4:27 AM, Chris Mi wrote: Hi Ilya, I think about your suggestion recently. But I'm still not very clear about the

[ovs-dev] [PATCH] ofproto-dpif-upcall: Improve concurrency by adjust flow-limit

OVS uses a set of gentle mechanisms in ofproto-dpif-upcall.c to control the number of flows in the datapath. However, this mechanism will make it difficult for the number of datapath flows to reach a high level in a short time. Through some tests, I found that the problem is that the flow-limit