Re: [ovs-dev] [PATCH v3] route-table: Filter route changes by interface.

On Tue, Mar 26, 2024 at 06:42:02PM +0100, 【外部账号】Ilya Maximets wrote: > On 3/24/24 13:16, Cheng Li wrote: > > When ovs host is also a kubernets node, pod creation/deletion may > > trigger route changes. As a result, ovs run route_table_reset(). > > As ovs do not care the kubernetes pod routes,

Re: [ovs-dev] [PATCH v2 1/5] ovsdb: raft: Avoid transferring leadership to unavailable servers.

On Tue, Mar 26, 2024 at 10:26 AM Ilya Maximets wrote: > > Current implementation of the leadership transfer just shoots the > leadership in the general direction of the first stable server in the > configuration. It doesn't check if the server was active recently or > even that the connection is

Re: [ovs-dev] [PATCH ovn v3 6/8] utilities/docker: Fix up container build.

On 3/25/24 05:09, Ales Musil wrote: On Mon, Mar 25, 2024 at 9:47 AM Dumitru Ceara > wrote: On 3/22/24 19:34, Mark Michelson wrote: > On 3/21/24 19:15, Dumitru Ceara wrote: >> Most of the steps were inaccurate.  Instead, use latest Ubuntu, use >>

Re: [ovs-dev] [PATCH v3] route-table: Filter route changes by interface.

On 3/24/24 13:16, Cheng Li wrote: > When ovs host is also a kubernets node, pod creation/deletion may > trigger route changes. As a result, ovs run route_table_reset(). > As ovs do not care the kubernetes pod routes, route_table_reset() > is not neccessary. > > Signed-off-by: Cheng Li > --- Hi,

[ovs-dev] [PATCH v2 5/5] ovsdb: raft: Fix inability to join after leadership change round trip.

Consider the following sequence of events: 1. Cluster with 2 nodes - A and B. A is a leader. 2. C connects to A and sends a join request. 3. A sends an append request to C. C is in CATCHUP phase for A. 4. A looses leadership to B. Sends join failure notification to C. 5. C sends append

[ovs-dev] [PATCH v2 3/5] ovsdb: raft: Fix permanent joining state on a cluster member.

Consider the following chain of events: 1. Have a cluster with 2 members - A and B. A is a leader. 2. C connects to A, sends a request to join the cluster. 3. A catches up C, creates an update for the 'servers' list and sends it to B and C to apply. This entry is not committed yet. 4.

[ovs-dev] [PATCH v2 4/5] ovsdb: raft: Fix assertion when 1-node cluster looses leadership.

Some of the failure tests can make a single-node cluster to loose leadership. In this case the next raft_run() will trigger election with a pre-vote enabled. This is causing an assertion when this server attempts to vote for itself. Fix that by not using pre-voting if there is only one server.

[ovs-dev] [PATCH v2 2/5] ovsdb: raft: Fix time intervals for multitasking while joining.

While joining, ovsdb-server may not wake up for a duration of a join timer, which is 1 second and is by default 3x larger than a heartbeat timer. This is causing unnecessary warnings from the cooperative multitasking module that thinks that we missed the heartbeat time by a lot. Use join timer

[ovs-dev] [PATCH v2 1/5] ovsdb: raft: Avoid transferring leadership to unavailable servers.

Current implementation of the leadership transfer just shoots the leadership in the general direction of the first stable server in the configuration. It doesn't check if the server was active recently or even that the connection is established. This may result in sending leadership to a

[ovs-dev] [PATCH v2 0/5] ovsdb: raft: Fixes for cluster joining state.

Issues discovered while working on: https://github.com/ovn-org/ovn/issues/235 Version 2: * Replaced the first patch with an implementation of leadership transfer to servers that applied the most changes. [Felix, Han] * Fixed some typos. [Han] * Added a MACRO for the join timeout

Re: [ovs-dev] [PATCH ovn v2 2/2] missed during test simplification.

Bleep bloop. Greetings Jacob Tanenbaum, 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: The subject summary should start with a capital. Subject: missed during test

[ovs-dev] [PATCH ovn v2 2/2] missed during test simplification.

the testing was changed to remove the hardcoded table numbers from flows. Two referances to table=28 where missed, this patch corrects that and adds the correct variable name to the flow. Fixes: f614335abca1 ("tests: Remove table numbers from "action parsing".") Signed-off-by: Jacob Tanenbaum

[ovs-dev] [PATCH ovn v2 1/2] Merge QoS logical pipelines.

currently there are 2 QoS pipelines for ingress (ls_in_qos_mark, ls_in_qos_meter) and egress (ls_out_qos_mark, ls_out_qos_meter). This is not necessary as there are no actions across the two pipelines that depend on each other. The two pipelines can be merged. Reported-at:

Re: [ovs-dev] [PATCH] ofproto-dpif-upcall: Don't mirror packets that aren't modified.

On Mon, Mar 25, 2024 at 10:04 PM Zhangweiwei wrote: > > The ethernet addresses of two ICMP request packets are indeed different . One > is original packet and the other is modified. It is an expected behavior > according to the code. > Actually, when a packet sent by port A is changed by flow

Re: [ovs-dev] [PATCH v13 5/6] tests: system-traffic: Add coverage for drop action.

On 3/22/24 15:34, Eelco Chaudron wrote: > > > On 22 Mar 2024, at 14:54, Eric Garver wrote: > >> Exercise the drop action in the datapath. This specific tests triggers >> an xlate_error. >> >> For the kernel datapath skb drop reasons can then be seen while this >> test runs. >> >> # perf trace

[ovs-dev] [PATCH ovn] tests: Add macro for checking flows after recompute.

The macro CHECK_FLOWS_AFTER_RECOMPUTE dumps the Openflows, then recomputes, then dumps again the Openflows, and finally compares both sets of flows. The test fails if flows are different. As of now, the macro cannot be used in all tests: many tests would fail as I+P does not properly remove flows

[ovs-dev] [PATCH ovn] automake: Make system tests dependent of ovn-macro.

So system testsuite will be recompiled when ovn-macro is changed. Signed-off-by: Xavier Simonart --- tests/automake.mk | 1 + 1 file changed, 1 insertion(+) diff --git a/tests/automake.mk b/tests/automake.mk index f6f0f0e33..1fdc89835 100644 --- a/tests/automake.mk +++ b/tests/automake.mk @@

[ovs-dev] [PATCH ovn v2 1/3] tests: Add macros to pause controller updates.

Such macros can then be used for instance to create condition where sb is seen as read-only by ovn-controller. Signed-off-by: Xavier Simonart --- tests/ovn-macros.at | 25 - 1 file changed, 24 insertions(+), 1 deletion(-) diff --git a/tests/ovn-macros.at

[ovs-dev] [PATCH ovn v2 2/3] pinctrl: Fix missing MAC_Bindings.

Pinctrl is responsible of creating MAC_Bindings on peer router datapaths. However, when sb was read-only, this did not happen. This caused the test "neighbor update on same HV" to fail in a flaky way. Signed-off-by: Xavier Simonart --- v2: - Fix userspace tests - Replace iptables by

[ovs-dev] [PATCH ovn v2 3/3] pinctrl: Fixed 100% cpu on ovs connection loss.

This issue is happening for instance when running test "ovn-controller - Chassis other_config". Signed-off-by: Xavier Simonart --- v2: Amend subject summary. Rebase on origin/main. --- controller/pinctrl.c | 15 --- 1 file changed, 8 insertions(+), 7 deletions(-) diff --git

Re: [ovs-dev] [PATCH ovn v2] acl-log: Properly log the "pass" verdict.

On Mon, Mar 25, 2024 at 6:48 PM Mark Michelson wrote: > The "pass" verdict was not explicitly defined in the list of verdicts > for ACL logging. This resulted in logs saying "Syntax error at `pass' > unknown verdict." > > This change adds the "pass" verdict explicitly so that it shows up as a >

Re: [ovs-dev] [PATCH ovn 2/2] missed during test simplification

On Mon, Mar 25, 2024 at 5:59 PM Jacob Tanenbaum wrote: Hi Jacob, thank you for the patch. Please address the 0-day bot concerns in v2. commit: 1807499 changed the testing to get the table numbers themselves. > We use the following tag to indicate that the commit is fixing some previous one:

Re: [ovs-dev] [PATCH ovn 1/2] Merge QoS logical pipelines

On Mon, Mar 25, 2024 at 5:59 PM Jacob Tanenbaum wrote: Hi Jacob, thank you for he patch, the subject should end with a dot, and there is one more thing down below. currently there are 2 QoS pipelines for ingress (ls_in_qos_mark, > ls_in_qos_meter) and egress (ls_out_qos_mark,

Re: [ovs-dev] [PATCH ovn] controller: Fix ofctrl memory usage underflow.

On Wed, Mar 20, 2024 at 12:48 PM Mark Michelson wrote: > > Thanks Ales, looks good to me. > > Acked-by: Mark Michelson > Thanks Ales and Mark. I applied to main and backported down to branch-23.06. Han > On 3/19/24 11:57, Ales Musil wrote: > > The memory usage would be increased for size of