0001 What: /sys/fs/ext4/<disk>/mb_stats
0002 Date: March 2008
0003 Contact: "Theodore Ts'o" <tytso@mit.edu>
0004 Description:
0005 Controls whether the multiblock allocator should
0006 collect statistics, which are shown during the unmount.
0007 1 means to collect statistics, 0 means not to collect
0008 statistics
0009
0010 What: /sys/fs/ext4/<disk>/mb_group_prealloc
0011 Date: March 2008
0012 Contact: "Theodore Ts'o" <tytso@mit.edu>
0013 Description:
0014 The multiblock allocator will round up allocation
0015 requests to a multiple of this tuning parameter if the
0016 stripe size is not set in the ext4 superblock
0017
0018 What: /sys/fs/ext4/<disk>/mb_max_to_scan
0019 Date: March 2008
0020 Contact: "Theodore Ts'o" <tytso@mit.edu>
0021 Description:
0022 The maximum number of extents the multiblock allocator
0023 will search to find the best extent
0024
0025 What: /sys/fs/ext4/<disk>/mb_min_to_scan
0026 Date: March 2008
0027 Contact: "Theodore Ts'o" <tytso@mit.edu>
0028 Description:
0029 The minimum number of extents the multiblock allocator
0030 will search to find the best extent
0031
0032 What: /sys/fs/ext4/<disk>/mb_order2_req
0033 Date: March 2008
0034 Contact: "Theodore Ts'o" <tytso@mit.edu>
0035 Description:
0036 Tuning parameter which controls the minimum size for
0037 requests (as a power of 2) where the buddy cache is
0038 used
0039
0040 What: /sys/fs/ext4/<disk>/mb_stream_req
0041 Date: March 2008
0042 Contact: "Theodore Ts'o" <tytso@mit.edu>
0043 Description:
0044 Files which have fewer blocks than this tunable
0045 parameter will have their blocks allocated out of a
0046 block group specific preallocation pool, so that small
0047 files are packed closely together. Each large file
0048 will have its blocks allocated out of its own unique
0049 preallocation pool.
0050
0051 What: /sys/fs/ext4/<disk>/inode_readahead_blks
0052 Date: March 2008
0053 Contact: "Theodore Ts'o" <tytso@mit.edu>
0054 Description:
0055 Tuning parameter which controls the maximum number of
0056 inode table blocks that ext4's inode table readahead
0057 algorithm will pre-read into the buffer cache
0058
0059 What: /sys/fs/ext4/<disk>/delayed_allocation_blocks
0060 Date: March 2008
0061 Contact: "Theodore Ts'o" <tytso@mit.edu>
0062 Description:
0063 This file is read-only and shows the number of blocks
0064 that are dirty in the page cache, but which do not
0065 have their location in the filesystem allocated yet.
0066
0067 What: /sys/fs/ext4/<disk>/lifetime_write_kbytes
0068 Date: March 2008
0069 Contact: "Theodore Ts'o" <tytso@mit.edu>
0070 Description:
0071 This file is read-only and shows the number of kilobytes
0072 of data that have been written to this filesystem since it was
0073 created.
0074
0075 What: /sys/fs/ext4/<disk>/session_write_kbytes
0076 Date: March 2008
0077 Contact: "Theodore Ts'o" <tytso@mit.edu>
0078 Description:
0079 This file is read-only and shows the number of
0080 kilobytes of data that have been written to this
0081 filesystem since it was mounted.
0082
0083 What: /sys/fs/ext4/<disk>/inode_goal
0084 Date: June 2008
0085 Contact: "Theodore Ts'o" <tytso@mit.edu>
0086 Description:
0087 Tuning parameter which (if non-zero) controls the goal
0088 inode used by the inode allocator in preference to
0089 all other allocation heuristics. This is intended for
0090 debugging use only, and should be 0 on production
0091 systems.
0092
0093 What: /sys/fs/ext4/<disk>/max_writeback_mb_bump
0094 Date: September 2009
0095 Contact: "Theodore Ts'o" <tytso@mit.edu>
0096 Description:
0097 The maximum number of megabytes the writeback code will
0098 try to write out before move on to another inode.
0099
0100 What: /sys/fs/ext4/<disk>/extent_max_zeroout_kb
0101 Date: August 2012
0102 Contact: "Theodore Ts'o" <tytso@mit.edu>
0103 Description:
0104 The maximum number of kilobytes which will be zeroed
0105 out in preference to creating a new uninitialized
0106 extent when manipulating an inode's extent tree. Note
0107 that using a larger value will increase the
0108 variability of time necessary to complete a random
0109 write operation (since a 4k random write might turn
0110 into a much larger write due to the zeroout
0111 operation).
0112
0113 What: /sys/fs/ext4/<disk>/journal_task
0114 Date: February 2019
0115 Contact: "Theodore Ts'o" <tytso@mit.edu>
0116 Description:
0117 This file is read-only and shows the pid of journal thread in
0118 current pid-namespace or 0 if task is unreachable.
