On 17/07/17 11:10, Numan Siddique wrote:


On Mon, Jul 17, 2017 at 1:10 PM, Eelco Chaudron <[email protected] <mailto:[email protected]>> wrote:

    On 14/07/17 18:59, Han Zhou wrote:



        On Fri, Jul 14, 2017 at 1:35 AM, Eelco Chaudron
        <[email protected] <mailto:[email protected]>
        <mailto:[email protected] <mailto:[email protected]>>> wrote:
        >
        > On 14/07/17 08:44, Han Zhou wrote:
        >>
        >> This reverts commit 67ac844b55d4c5f6bbfa01773c82b3d6d8b62131.
        >>
        >> The commit introduced a problem that "File exists" will be
        reported
        >> when trying to open br0.
        >>
        >> The operation that adds eth0 to br0 while moving IP address
        from
        >> eth0 to bridge internal interface br0 reproduces this issue.
        >>
        >> $ ip a del <ip> dev eth0; ip a add <ip> dev br0; ovs-vsctl
        add-port br0 eth0
        >> $ ovs-dpctl show
        >> ...
        >> port 1: br0 (internal: open failed (File exists))
        >> ...
        >>
        >> At this point restarting OVS will result in connection lost
        for the
        >> node.
        >>
        >> Reverting the change fixes the problem. Since adding
        physical interface
        >> to OVS bridge is quite normal operation, the problem is
        more severe
        >> than the original problem fixed by commit 67ac844, so
        revert this
        >> before a better fix is found for the original problem.
        >
        > See also this thread:
        >
        >
        https://mail.openvswitch.org/pipermail/ovs-dev/2017-July/335428.html
        <https://mail.openvswitch.org/pipermail/ovs-dev/2017-July/335428.html>
        >
        > Will be sending out the patch later today.
        >
        > //Eelco*
        >
        Thanks Eelco! Too bad I didn't notice the thread yesterday,
        otherwise I would not spend hours chasing the problem :(
        Since the problem is critical (e.g. installing release v2.7.1
        on any existing hypervisors would break connection), do you
        think it is better to revert the commit until we are confident
        with the new patch?

    Assuming we can get the patch in before the next release, there
    should be no need to revert back. But if its common practice for
    OVS to do it right away I have no issue with it.


For tripleo, we need OVS 2.7.1 for OVN DB HA support, but we can't take v2.7.1 for this issue. I think we should atleast revert the patch for 2.7 branch and have a new version 2.7.2.

If we are having a 2.7.2 release specially for this it makes sense to revert it.

In addition can you test/apply the updated patch, https://mail.openvswitch.org/pipermail/ovs-dev/2017-July/335560.html as I was hitting the issue with the persistant ports setup, it would be good to see your case is also fixed.

Thanks,

Eelco

_______________________________________________
dev mailing list
[email protected]
https://mail.openvswitch.org/mailman/listinfo/ovs-dev

Reply via email to