[ovs-dev] [lacp 03/10] bridge: Pull monitoring out of bonding logic.

Ethan Jackson ethan at nicira.com
Thu Mar 17 19:40:17 UTC 2011


Looking at it more carefully now, I realize why I put it in bridge
reconfigure.   In port_reconfigure, interfaces do not necessarily have
their netdevs initialized yet.  Monitoring setup has to happen after
they are initialized in bridge reconfigure.

On Wed, Mar 16, 2011 at 10:58 AM, Ethan Jackson <ethan at nicira.com> wrote:
> Artifact of monitoring being done as part of bonding.  I added the
> monitors neer port_update_bonding.  No particular reason other than
> history. I will go ahead and change it.
>
> Ethan
>
> On Wed, Mar 16, 2011 at 9:24 AM, Ben Pfaff <blp at nicira.com> wrote:
>> On Tue, Mar 15, 2011 at 04:03:16PM -0700, Ethan Jackson wrote:
>>> The monitoring logic and bonding code are unrelated.  This commit
>>> pulls the monitoring logic out.  As a result all interfaces, not
>>> just those participating in bonds, are monitored.  This will be
>>> required to run LACP on non-bonded ports.
>>>
>>> Also removes the miimon flag from the port structure.
>>
>> This looks OK to me, but one bit stuck out to me as not fitting in
>> well.  The patch adds code to bridge_reconfigure() to call
>> netdev_monitor_add() for each interface that should be monitored.  I'm
>> curious about why it goes in there and not, for example, in
>> port_reconfigure(), which to my mind would be a more natural place
>> (since that's the code that creates or destroys the monitor itself).
>>
>> Thanks,
>>
>> Ben.
>>
>



More information about the dev mailing list