> On Nov. 11, 2012, 8:03 p.m., Ben Mahler wrote: > > src/slave/slave.cpp, line 414 > > <https://reviews.apache.org/r/8009/diff/1/?file=188446#file188446line414> > > > > We only attach the executor work directories, so there should be no > > need to detach here and immediately below. Am I missing something?
So, I'm thinking about the case when a slave restarts. Currently, the slave gets a new slave id and hence a new directory /tmp/mesos/slaves/<slave-id>. The old slave directory (and the corresponding executor directories) get gc'ed here. What happens to the webui in this case? Since, we never attached the old executor directories after the slave restarts, it is ok i suppose? - Vinod ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/8009/#review13339 ----------------------------------------------------------- On Nov. 11, 2012, 2:27 a.m., Ben Mahler wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/8009/ > ----------------------------------------------------------- > > (Updated Nov. 11, 2012, 2:27 a.m.) > > > Review request for mesos, Benjamin Hindman and Vinod Kone. > > > Description > ------- > > The attached directories map inside files.cpp grows slowly forever. > This remedies the "leak". > > > Diffs > ----- > > src/slave/gc.hpp dc421e03cad9acb994adf794ffa59f6b8636fc8d > src/slave/slave.hpp bbba4404e9e2b1ff1e246f017cdad704438973ba > src/slave/slave.cpp 7deb4574943aae4cfc5da5d6b3f600042686975f > src/tests/gc_tests.cpp 6b3ba02685e54cc772df6b3c929355cb47f54d52 > > Diff: https://reviews.apache.org/r/8009/diff/ > > > Testing > ------- > > Added some tests to gc_tests.cpp > > > Thanks, > > Ben Mahler > >
