Orit Wasserman <owass...@redhat.com> wrote: > On 02/01/2013 02:32 PM, Juan Quintela wrote: >> While we are sleeping we are not sending, so we should not use that >> time to estimate our bandwidth. > Juan, > Maybe I missing something here but the sleep time is cause by the fact > we limit the migration bandwidth. So the available bandwidth should take > it into consideration? we still cape the bandwidth in the last stage. > Regards,
No, in the last stage we sent as far as we can. If last_round == true, we dont' enter the if with the g_usleep(). Later, Juan.
