4 An 'obd_statfs' structure conveys file-system-wide information of a
5 given target (MDT or OST).
34 Most of the fields correspond directly to the 'struct statfs' fields
35 returned by a 'statfs()' system call and have the same meaning. The
36 values are for the target in question (MDT or OST).
38 The 'os_type' field gives the kind of file system used for the
41 .File system types ('os_type') for Lustre targets
45 | EXT4_SUPER_MAGIC (ldiskfs) | 0xEF53
46 | ZFS_SUPER_MAGIC | 0x2fc12fc1
49 The 'os_blocks' field is the total number of blocks in the target, in
52 The 'os_bfree' field is the number of blocks not currently in use.
54 The 'os_bavail' is the number of blocks available to be allocated to
55 new files. The number of available blocks is typically lower than
56 the number of free blocks due to <<grant>> and blocks reserved for
57 target internal usage such as metadata.
59 The 'os_files' field is the total number of files on the target, both
60 allocated and free. For some OSD types this is a static number, and
61 for others this is dynamic based on os_ffree and the amount of free
64 The 'os_ffree' field is the current number of files that could be
65 created based on current usage. For some OSD types this is a function
66 of the free space in the device and may increase or decrease as free
69 The 'os_fsid' is intended to be the target backing device UUID in
70 ASCII format. The current osd-ldiskfs and osd-zfs implementations
71 do not fill this field in.
73 The 'os_bsize' field is the block size in bytes. This is for computing
74 the total, free, and available space in combination with 'os_blocks',
75 'os_bfree', and 'os_bavail' respectively. It does not necessarily
76 represent the minimum or optimal IO size.
78 The 'os_namelen' field gives the maximum name length (in bytes) for
81 The 'os_maxbytes' field is the maximum size of a single object
82 (i.e. the maximum byte offset that can be written to). This is the
83 same value as the 'ocd_maxbytes' field of the 'obd_connect_data'
86 The 'os_state' field encodes the status of the OSD. It can be:
91 | os_state flag | value
92 | OS_STATE_DEGRADED | 0x1
93 | OS_STATE_READONLY | 0x2
96 In normal operation the 'os_state' value is returned as 0x0. If the
97 OSD has a RAID configuration that is degraded or
98 rebuilding the state is returned with the OS_STATE_DEGRADED (0x1) flag
99 set. If the file system has been set to read-only, either manually at
100 mount or automatcially due to detected corruption of the underlying
101 target filesystem, then 'os_state' is returned with OS_STATE_READONLY (0x2)
102 set, for example if it was explicitly mounted read-only, or
103 corruption has been detected at runtime in the backing filesystem.
105 The 'os_fprecreated' field counts the number of pre-created objects
106 available on an OST. The 'os_fprecreated' value counts as "used"
107 inodes, so it reduces 'os_ffree'. Lustre places a hard limit of
108 2*OST_MAX_PRECREATE (2*20000) on the number of precreates it will
109 allow for an OST. There are currently no precreated objects on an MDT
110 so it is just 0 in that case.