> On June 7, 2016, 2:09 a.m., Guangya Liu wrote: > > src/tests/environment.cpp, lines 492-495 > > <https://reviews.apache.org/r/47396/diff/4/?file=1407798#file1407798line492> > > > > I think that here also needs to be updated to use your new function > > `fs::aufs::supported()` to check `aufs`? > > Jie Yu wrote: > Hum... this looks weird to me. I am wondering if we can make > `fs::supported` understand the sutble details related to overlayfs? Can you > guys follow up with a patch. Basically, here, it should be `Try<bool> check = > fs::supported(fsname);` and inside `fs::supported`, we check special case for > overlayfs. > > Guangya Liu wrote: > I'm now thinking that we may not need the API of `fs::aufs::supported`, > the reason that we introduced `fs::overlay::supported` is becuase of keyword > issue, the overlay fs keyword in `/proc/filesystems` can be `overlay` or > `overlayfs`. > > But for `aufs`, I think we do not need to introduce the new API of > `fs::aufs::supported` but call `fs::supported` directly, as the API of > `fs::aufs::supported` is just a wrapper of `fs::supported`. Comments? > > Jie Yu wrote: > Yeah, I agree with you Guangya. I think we should make > `fs::supported(fsname)` understand both overlayfs and overlay and get rid of > all fs::xxx::supported logics. Do you have time to follow up with a patch? > > Guangya Liu wrote: > Hi Yu Jie, it may be difficult to make `fs::supported(fsname)` to > understand both overlayfs and overlay, so my thinking is that we only keep > `fs::overlay::supported` but remove the `fs::aufs::supported`, comments?
Why it's difficult to make fs::supported(fsname) to understand both overlayfs and overlay? - Jie ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/47396/#review136392 ----------------------------------------------------------- On June 6, 2016, 11:57 p.m., Shuai Lin wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/47396/ > ----------------------------------------------------------- > > (Updated June 6, 2016, 11:57 p.m.) > > > Review request for mesos, Gilbert Song, Guangya Liu, and Jie Yu. > > > Bugs: MESOS-4672 > https://issues.apache.org/jira/browse/MESOS-4672 > > > Repository: mesos > > > Description > ------- > > Added aufs provisioning backend. > > > Diffs > ----- > > src/Makefile.am a08ea407d631f6ae56ac36b122bfdf0e849e8b56 > src/slave/containerizer/mesos/provisioner/backend.cpp > b2a20b7c08fa790da09ba05b725248e42b8d3bc4 > src/slave/containerizer/mesos/provisioner/backends/aufs.hpp PRE-CREATION > src/slave/containerizer/mesos/provisioner/backends/aufs.cpp PRE-CREATION > src/tests/containerizer/provisioner_backend_tests.cpp > bc04760c9504ea55cd4bb72c7e9012e43a5911aa > src/tests/environment.cpp 849e9ce05864f6ec1a736dfc1a7a31d2364c84bf > > Diff: https://reviews.apache.org/r/47396/diff/ > > > Testing > ------- > > - Make check on ubuntu 14.04 64bit > - Test manually: start slave with aufs backend, and write a simple script > that launches tasks with alpine/ubuntu images > > > Thanks, > > Shuai Lin > >