On 9/2/22 10:52, Eelco Chaudron wrote:
> 
> 
> On 2 Sep 2022, at 10:46, David Marchand wrote:
> 
>> On Fri, Sep 2, 2022 at 10:35 AM Eelco Chaudron <[email protected]> wrote:
>>> On 1 Sep 2022, at 18:02, Eelco Chaudron wrote:
>>>
>>>> Using [NETNS|OVS]_DAEMONIZE will start tcpdump in the background,
>>>> and it will also make sure it gets killed in corner cases.
>>>>
>>>> For the check_pkt_len tests, we also kill tcpdump between individual
>>>> tests in the same test case to avoid confusion when analyzing results.
>>>> This also required some changes to the awk expressions, as an extra
>>>> newline is added to the output when tcpdump gets stopped.
>>>>
>>>> Fixes: 02dabb21f243 ("tests: Add check_pkt_len action test to 
>>>> system-offload-traffic.")
>>>> Suggested-by: David Marchand <[email protected]>
>>>> Signed-off-by: Eelco Chaudron <[email protected]>
>>>
>>> Looks like the 0-day Robot failure is not related to this patch.
>>>
>>> I’ve run this through GitHub testing and it’s fine:
>>>
>>> https://github.com/chaudron/ovs/actions?query=
>>
>> I once hit this BFD failure in GHA too but could not reproduce it on my side.
> 
> Guess it’s a failure that happens randomly (probably timing based), also the 
> failing test has not been modified by this patch.

'BFD decay' is a known flaky test.  It is time based.
I have a potential fix for it, but I'm not 100% sure
that it fixes the problem completely, since I'm also
not reproducing the problem 100% of the time.  At least,
I don't see the failure on my laptop with the change.
I'll try to brush it up and post.

Bets regards, Ilya Maximets.
_______________________________________________
dev mailing list
[email protected]
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Reply via email to