cgroup-v2.rst (382625d0d4325fb14a29444eb8dce8dcc2eb9b51) | cgroup-v2.rst (2324d50d051ec0f14a548e78554fb02513d6dcef) |
---|---|
1================ 2Control Group v2 3================ 4 5:Date: October, 2015 6:Author: Tejun Heo <tj@kernel.org> 7 8This is the authoritative documentation on the design, interface and --- 1669 unchanged lines hidden (view full) --- 1678implements control of page cache writeback IOs. The memory controller 1679defines the memory domain that dirty memory ratio is calculated and 1680maintained for and the io controller defines the io domain which 1681writes out dirty pages for the memory domain. Both system-wide and 1682per-cgroup dirty memory states are examined and the more restrictive 1683of the two is enforced. 1684 1685cgroup writeback requires explicit support from the underlying | 1================ 2Control Group v2 3================ 4 5:Date: October, 2015 6:Author: Tejun Heo <tj@kernel.org> 7 8This is the authoritative documentation on the design, interface and --- 1669 unchanged lines hidden (view full) --- 1678implements control of page cache writeback IOs. The memory controller 1679defines the memory domain that dirty memory ratio is calculated and 1680maintained for and the io controller defines the io domain which 1681writes out dirty pages for the memory domain. Both system-wide and 1682per-cgroup dirty memory states are examined and the more restrictive 1683of the two is enforced. 1684 1685cgroup writeback requires explicit support from the underlying |
1686filesystem. Currently, cgroup writeback is implemented on ext2, ext4 1687and btrfs. On other filesystems, all writeback IOs are attributed to 1688the root cgroup. | 1686filesystem. Currently, cgroup writeback is implemented on ext2, ext4, 1687btrfs, f2fs, and xfs. On other filesystems, all writeback IOs are 1688attributed to the root cgroup. |
1689 1690There are inherent differences in memory and writeback management 1691which affects how cgroup ownership is tracked. Memory is tracked per 1692page while writeback per inode. For the purpose of writeback, an 1693inode is assigned to a cgroup and all IO requests to write dirty pages 1694from the inode are attributed to that cgroup. 1695 1696As cgroup ownership for memory is tracked per page, there can be pages --- 340 unchanged lines hidden (view full) --- 2037An example of BPF_CGROUP_DEVICE program may be found in the kernel 2038source tree in the tools/testing/selftests/bpf/dev_cgroup.c file. 2039 2040 2041RDMA 2042---- 2043 2044The "rdma" controller regulates the distribution and accounting of | 1689 1690There are inherent differences in memory and writeback management 1691which affects how cgroup ownership is tracked. Memory is tracked per 1692page while writeback per inode. For the purpose of writeback, an 1693inode is assigned to a cgroup and all IO requests to write dirty pages 1694from the inode are attributed to that cgroup. 1695 1696As cgroup ownership for memory is tracked per page, there can be pages --- 340 unchanged lines hidden (view full) --- 2037An example of BPF_CGROUP_DEVICE program may be found in the kernel 2038source tree in the tools/testing/selftests/bpf/dev_cgroup.c file. 2039 2040 2041RDMA 2042---- 2043 2044The "rdma" controller regulates the distribution and accounting of |
2045of RDMA resources. | 2045RDMA resources. |
2046 2047RDMA Interface Files 2048~~~~~~~~~~~~~~~~~~~~ 2049 2050 rdma.max 2051 A readwrite nested-keyed file that exists for all the cgroups 2052 except root that describes current configured resource limit 2053 for a RDMA/IB device. --- 546 unchanged lines hidden --- | 2046 2047RDMA Interface Files 2048~~~~~~~~~~~~~~~~~~~~ 2049 2050 rdma.max 2051 A readwrite nested-keyed file that exists for all the cgroups 2052 except root that describes current configured resource limit 2053 for a RDMA/IB device. --- 546 unchanged lines hidden --- |