> On March 26, 2016, 12:37 a.m., Jie Yu wrote: > > src/slave/slave.cpp, line 2375 > > <https://reviews.apache.org/r/45121/diff/1/?file=1308847#file1308847line2375> > > > > I would suggest we don't use CHECK here. We can just LOG(ERROR) if the > > deletion fails. Given the TODO above, we could leak disk space anyway.
I used `CHECK_SOME` because we use the same error-handling strategy when creating the persistent volume fails, or if the slave fails to checkpoint resources to disk successfully. Are you sure we want to handle `rmdir` errors differently? - Neil ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/45121/#review125492 ----------------------------------------------------------- On March 21, 2016, 6:44 p.m., Neil Conway wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/45121/ > ----------------------------------------------------------- > > (Updated March 21, 2016, 6:44 p.m.) > > > Review request for mesos, Jie Yu and Joris Van Remoortere. > > > Bugs: MESOS-2408 > https://issues.apache.org/jira/browse/MESOS-2408 > > > Repository: mesos > > > Description > ------- > > Prior to this commit, destroying a persistent volume would remove > the Mesos-level metadata about the volume, but wouldn't destroy > any of the volume's filesystem content. We now remove the volume > from the slave's filesystem, essentially via "rm -r". > > > Diffs > ----- > > src/slave/slave.cpp 840534ff0687e82ed063c386e36bbabada230697 > > Diff: https://reviews.apache.org/r/45121/diff/ > > > Testing > ------- > > make check > > > Thanks, > > Neil Conway > >
