----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/68401/#review208778 -----------------------------------------------------------
src/slave/containerizer/mesos/isolators/xfs/disk.hpp Lines 83 (patched) <https://reviews.apache.org/r/68401/#comment292961> Nit: `const`? src/slave/containerizer/mesos/isolators/xfs/disk.hpp Lines 87 (patched) <https://reviews.apache.org/r/68401/#comment292958> Nit: We can use parameter names without underscores here - there's no shadowing. Also, `explicit`? Unless we want to use `= {dir, projid}` for it :) src/slave/containerizer/mesos/isolators/xfs/disk.cpp Lines 342-345 (patched) <https://reviews.apache.org/r/68401/#comment292968> Follow up to the comment below: if multiple volumes can have the same project ID `project_ids_free` metric will be overdecremented here. src/slave/containerizer/mesos/isolators/xfs/disk.cpp Lines 347-348 (patched) <https://reviews.apache.org/r/68401/#comment292967> Are we checking this in case multiple volumes have the same project ID? If that can be true, later we may falsely free a project ID that is still in use by the volume that we ignored here. Do we elect to ignore such misconfigurations? src/slave/containerizer/mesos/isolators/xfs/disk.cpp Lines 484-487 (patched) <https://reviews.apache.org/r/68401/#comment292971> IMHO this can justify keeping dedicated `directory`, `projectId` and `quota` for container sandbox in `Info` :) But this is just my taste. src/slave/containerizer/mesos/isolators/xfs/disk.cpp Lines 529 (patched) <https://reviews.apache.org/r/68401/#comment292972> Add volume ID? src/slave/containerizer/mesos/isolators/xfs/disk.cpp Line 416 (original), 585-587 (patched) <https://reviews.apache.org/r/68401/#comment292975> Funny how in a case where a volume is shared by multiple containers we can kill all of them because of a single offender. Just thinking out loud. Probably worth noting that in the docs. src/slave/containerizer/mesos/isolators/xfs/disk.cpp Lines 421-422 (original), 592-593 (patched) <https://reviews.apache.org/r/68401/#comment292973> Log which directory since now we check several of them? src/slave/containerizer/mesos/isolators/xfs/disk.cpp Lines 438-440 (original), 609-611 (patched) <https://reviews.apache.org/r/68401/#comment292974> Add which quota, i.e. sandbox/volume? - Ilya Pronin On Sept. 10, 2018, 2:34 p.m., James Peach wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/68401/ > ----------------------------------------------------------- > > (Updated Sept. 10, 2018, 2:34 p.m.) > > > Review request for mesos, Chun-Hung Hsiao, Ilya Pronin, Jie Yu, Joseph Wu, > and Jiang Yan Xu. > > > Bugs: MESOS-5158 > https://issues.apache.org/jira/browse/MESOS-5158 > > > Repository: mesos > > > Description > ------- > > Added persistent volume support to the `disk/xfs` isolator. This > implementation largely tracks the `disk/du` implementation in that > we now keep a map of paths in each container info structure. We now > defer quota clean up to project ID reclaimation time so that we can > use the same mechanism for sandbox and persistent volume paths. > > We explicitly exclude mount disks from XFS project quotas, but we still > track them so that we can correctly publish their usage information in > the container `DiskStatistics` message. This means that mount disks are > not required to be XFS filesystems or have project quotas configured. > > > Diffs > ----- > > src/slave/containerizer/mesos/isolators/xfs/disk.hpp > 38c467b47cb7c04803b0709b8239458fb26abb61 > src/slave/containerizer/mesos/isolators/xfs/disk.cpp > 783da0407528c044035d18cc59a744353921d64c > src/tests/containerizer/xfs_quota_tests.cpp > 2b3a2e25f5075357f090d47826698e7bb6fdf969 > > > Diff: https://reviews.apache.org/r/68401/diff/4/ > > > Testing > ------- > > sudo make check (Fedora 28) > > > Thanks, > > James Peach > >
