[ovs-dev] [PATCH v2] FAQ: Add an entry about reconfiguration

[Ben Pfaff]

On Fri, Sep 19, 2014 at 12:24:38AM +0900, YAMAMOTO Takashi wrote:
> It seems that the behaviour is not so intuitive.
> cf. https://bugs.launchpad.net/neutron/+bug/1346861
> 
> Signed-off-by: YAMAMOTO Takashi <yamamoto at valinux.co.jp>

I am not sure that I understand the bug report there.  It might be
reporting an actual bug in OVS.

The goal of ovs-vswitchd regarding the database is to make sure that
the state of the system is kept up-to-date with whatever is in the
database.  Maybe that bug report is saying, "If I have a port, and
then I del-port/add-port that in a single transaction, ovs-vswitchd
does not actually delete a port and then readd it at the datapath
level."  If it is saying that, then it is correct.  But that could
also happen if you use multiple transactions, because when it is busy
ovs-vswitchd might "miss" some of the intermediate transactions and
just implement the overall effect.

The reason that two ovs-vsctl calls always deletes and readds a port
is a little different: the first ovs-vsctl waits for its transaction
to take effect before executing.  If you use "--no-wait", then you
just have two bare transactions and won't get the behavior of a
del-port followed by an add-port 100% of the time.