----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/58869/#review173454 -----------------------------------------------------------
Ship it! This change means that the cleanup of a container using docker volumes in the containerizer can be up to 20 mins. - Gilbert Song On April 28, 2017, 5:39 p.m., Jie Yu wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/58869/ > ----------------------------------------------------------- > > (Updated April 28, 2017, 5:39 p.m.) > > > Review request for mesos and Gilbert Song. > > > Bugs: MESOS-7439 > https://issues.apache.org/jira/browse/MESOS-7439 > > > Repository: mesos > > > Description > ------- > > A volume detach operation can be very time consuming in some cloud > environment. In extreme cases, the detach can take up to 20 minutes. > Since 'Unmount' in Docker Volume Driver Interface (DVDI) is not > idempotent, the container orchestrator (i.e., Mesos) cannot safely > retry the operation. As a result, we don't have much flexibility here. > This patch increases the timeout value hoping the operation will > finish within this new timeout value in most of the cases. > > > Diffs > ----- > > src/slave/containerizer/mesos/isolators/docker/volume/driver.cpp > 18a8066b064835cbd7a2cef6d1a43462bdd5e87c > > > Diff: https://reviews.apache.org/r/58869/diff/1/ > > > Testing > ------- > > make check > > > Thanks, > > Jie Yu > >
