Documentation/ABI/testing/sysfs-block

   1 What:           /sys/block/<disk>/stat
   2 Date:           February 2008
   3 Contact:        Jerome Marchand <jmarchan@redhat.com>
   4 Description:
   5                 The /sys/block/<disk>/stat files displays the I/O
   6                 statistics of disk <disk>. They contain 11 fields:
   7                  1 - reads completed successfully
   8                  2 - reads merged
   9                  3 - sectors read
  10                  4 - time spent reading (ms)
  11                  5 - writes completed
  12                  6 - writes merged
  13                  7 - sectors written
  14                  8 - time spent writing (ms)
  15                  9 - I/Os currently in progress
  16                 10 - time spent doing I/Os (ms)
  17                 11 - weighted time spent doing I/Os (ms)
  18                 For more details refer Documentation/iostats.txt
  19
  20
  21 What:           /sys/block/<disk>/<part>/stat
  22 Date:           February 2008
  23 Contact:        Jerome Marchand <jmarchan@redhat.com>
  24 Description:
  25                 The /sys/block/<disk>/<part>/stat files display the
  26                 I/O statistics of partition <part>. The format is the
  27                 same as the above-written /sys/block/<disk>/stat
  28                 format.
  29
  30
  31 What:           /sys/block/<disk>/integrity/format
  32 Date:           June 2008
  33 Contact:        Martin K. Petersen <martin.petersen@oracle.com>
  34 Description:
  35                 Metadata format for integrity capable block device.
  36                 E.g. T10-DIF-TYPE1-CRC.
  37
  38
  39 What:           /sys/block/<disk>/integrity/read_verify
  40 Date:           June 2008
  41 Contact:        Martin K. Petersen <martin.petersen@oracle.com>
  42 Description:
  43                 Indicates whether the block layer should verify the
  44                 integrity of read requests serviced by devices that
  45                 support sending integrity metadata.
  46
  47
  48 What:           /sys/block/<disk>/integrity/tag_size
  49 Date:           June 2008
  50 Contact:        Martin K. Petersen <martin.petersen@oracle.com>
  51 Description:
  52                 Number of bytes of integrity tag space available per
  53                 512 bytes of data.
  54
  55
  56 What:           /sys/block/<disk>/integrity/device_is_integrity_capable
  57 Date:           July 2014
  58 Contact:        Martin K. Petersen <martin.petersen@oracle.com>
  59 Description:
  60                 Indicates whether a storage device is capable of storing
  61                 integrity metadata. Set if the device is T10 PI-capable.
  62
  63 What:           /sys/block/<disk>/integrity/protection_interval_bytes
  64 Date:           July 2015
  65 Contact:        Martin K. Petersen <martin.petersen@oracle.com>
  66 Description:
  67                 Describes the number of data bytes which are protected
  68                 by one integrity tuple. Typically the device's logical
  69                 block size.
  70
  71 What:           /sys/block/<disk>/integrity/write_generate
  72 Date:           June 2008
  73 Contact:        Martin K. Petersen <martin.petersen@oracle.com>
  74 Description:
  75                 Indicates whether the block layer should automatically
  76                 generate checksums for write requests bound for
  77                 devices that support receiving integrity metadata.
  78
  79 What:           /sys/block/<disk>/alignment_offset
  80 Date:           April 2009
  81 Contact:        Martin K. Petersen <martin.petersen@oracle.com>
  82 Description:
  83                 Storage devices may report a physical block size that is
  84                 bigger than the logical block size (for instance a drive
  85                 with 4KB physical sectors exposing 512-byte logical
  86                 blocks to the operating system).  This parameter
  87                 indicates how many bytes the beginning of the device is
  88                 offset from the disk's natural alignment.
  89
  90 What:           /sys/block/<disk>/<partition>/alignment_offset
  91 Date:           April 2009
  92 Contact:        Martin K. Petersen <martin.petersen@oracle.com>
  93 Description:
  94                 Storage devices may report a physical block size that is
  95                 bigger than the logical block size (for instance a drive
  96                 with 4KB physical sectors exposing 512-byte logical
  97                 blocks to the operating system).  This parameter
  98                 indicates how many bytes the beginning of the partition
  99                 is offset from the disk's natural alignment.
 100
 101 What:           /sys/block/<disk>/queue/logical_block_size
 102 Date:           May 2009
 103 Contact:        Martin K. Petersen <martin.petersen@oracle.com>
 104 Description:
 105                 This is the smallest unit the storage device can
 106                 address.  It is typically 512 bytes.
 107
 108 What:           /sys/block/<disk>/queue/physical_block_size
 109 Date:           May 2009
 110 Contact:        Martin K. Petersen <martin.petersen@oracle.com>
 111 Description:
 112                 This is the smallest unit a physical storage device can
 113                 write atomically.  It is usually the same as the logical
 114                 block size but may be bigger.  One example is SATA
 115                 drives with 4KB sectors that expose a 512-byte logical
 116                 block size to the operating system.  For stacked block
 117                 devices the physical_block_size variable contains the
 118                 maximum physical_block_size of the component devices.
 119
 120 What:           /sys/block/<disk>/queue/minimum_io_size
 121 Date:           April 2009
 122 Contact:        Martin K. Petersen <martin.petersen@oracle.com>
 123 Description:
 124                 Storage devices may report a granularity or preferred
 125                 minimum I/O size which is the smallest request the
 126                 device can perform without incurring a performance
 127                 penalty.  For disk drives this is often the physical
 128                 block size.  For RAID arrays it is often the stripe
 129                 chunk size.  A properly aligned multiple of
 130                 minimum_io_size is the preferred request size for
 131                 workloads where a high number of I/O operations is
 132                 desired.
 133
 134 What:           /sys/block/<disk>/queue/optimal_io_size
 135 Date:           April 2009
 136 Contact:        Martin K. Petersen <martin.petersen@oracle.com>
 137 Description:
 138                 Storage devices may report an optimal I/O size, which is
 139                 the device's preferred unit for sustained I/O.  This is
 140                 rarely reported for disk drives.  For RAID arrays it is
 141                 usually the stripe width or the internal track size.  A
 142                 properly aligned multiple of optimal_io_size is the
 143                 preferred request size for workloads where sustained
 144                 throughput is desired.  If no optimal I/O size is
 145                 reported this file contains 0.
 146
 147 What:           /sys/block/<disk>/queue/nomerges
 148 Date:           January 2010
 149 Contact:
 150 Description:
 151                 Standard I/O elevator operations include attempts to
 152                 merge contiguous I/Os. For known random I/O loads these
 153                 attempts will always fail and result in extra cycles
 154                 being spent in the kernel. This allows one to turn off
 155                 this behavior on one of two ways: When set to 1, complex
 156                 merge checks are disabled, but the simple one-shot merges
 157                 with the previous I/O request are enabled. When set to 2,
 158                 all merge tries are disabled. The default value is 0 -
 159                 which enables all types of merge tries.
 160
 161 What:           /sys/block/<disk>/discard_alignment
 162 Date:           May 2011
 163 Contact:        Martin K. Petersen <martin.petersen@oracle.com>
 164 Description:
 165                 Devices that support discard functionality may
 166                 internally allocate space in units that are bigger than
 167                 the exported logical block size. The discard_alignment
 168                 parameter indicates how many bytes the beginning of the
 169                 device is offset from the internal allocation unit's
 170                 natural alignment.
 171
 172 What:           /sys/block/<disk>/<partition>/discard_alignment
 173 Date:           May 2011
 174 Contact:        Martin K. Petersen <martin.petersen@oracle.com>
 175 Description:
 176                 Devices that support discard functionality may
 177                 internally allocate space in units that are bigger than
 178                 the exported logical block size. The discard_alignment
 179                 parameter indicates how many bytes the beginning of the
 180                 partition is offset from the internal allocation unit's
 181                 natural alignment.
 182
 183 What:           /sys/block/<disk>/queue/discard_granularity
 184 Date:           May 2011
 185 Contact:        Martin K. Petersen <martin.petersen@oracle.com>
 186 Description:
 187                 Devices that support discard functionality may
 188                 internally allocate space using units that are bigger
 189                 than the logical block size. The discard_granularity
 190                 parameter indicates the size of the internal allocation
 191                 unit in bytes if reported by the device. Otherwise the
 192                 discard_granularity will be set to match the device's
 193                 physical block size. A discard_granularity of 0 means
 194                 that the device does not support discard functionality.
 195
 196 What:           /sys/block/<disk>/queue/discard_max_bytes
 197 Date:           May 2011
 198 Contact:        Martin K. Petersen <martin.petersen@oracle.com>
 199 Description:
 200                 Devices that support discard functionality may have
 201                 internal limits on the number of bytes that can be
 202                 trimmed or unmapped in a single operation. Some storage
 203                 protocols also have inherent limits on the number of
 204                 blocks that can be described in a single command. The
 205                 discard_max_bytes parameter is set by the device driver
 206                 to the maximum number of bytes that can be discarded in
 207                 a single operation. Discard requests issued to the
 208                 device must not exceed this limit. A discard_max_bytes
 209                 value of 0 means that the device does not support
 210                 discard functionality.
 211
 212 What:           /sys/block/<disk>/queue/discard_zeroes_data
 213 Date:           May 2011
 214 Contact:        Martin K. Petersen <martin.petersen@oracle.com>
 215 Description:
 216                 Devices that support discard functionality may return
 217                 stale or random data when a previously discarded block
 218                 is read back. This can cause problems if the filesystem
 219                 expects discarded blocks to be explicitly cleared. If a
 220                 device reports that it deterministically returns zeroes
 221                 when a discarded area is read the discard_zeroes_data
 222                 parameter will be set to one. Otherwise it will be 0 and
 223                 the result of reading a discarded area is undefined.
 224
 225 What:           /sys/block/<disk>/queue/write_same_max_bytes
 226 Date:           January 2012
 227 Contact:        Martin K. Petersen <martin.petersen@oracle.com>
 228 Description:
 229                 Some devices support a write same operation in which a
 230                 single data block can be written to a range of several
 231                 contiguous blocks on storage. This can be used to wipe
 232                 areas on disk or to initialize drives in a RAID
 233                 configuration. write_same_max_bytes indicates how many
 234                 bytes can be written in a single write same command. If
 235                 write_same_max_bytes is 0, write same is not supported
 236                 by the device.
 237
 238 What:           /sys/block/<disk>/queue/write_zeroes_max_bytes
 239 Date:           November 2016
 240 Contact:        Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
 241 Description:
 242                 Devices that support write zeroes operation in which a
 243                 single request can be issued to zero out the range of
 244                 contiguous blocks on storage without having any payload
 245                 in the request. This can be used to optimize writing zeroes
 246                 to the devices. write_zeroes_max_bytes indicates how many
 247                 bytes can be written in a single write zeroes command. If
 248                 write_zeroes_max_bytes is 0, write zeroes is not supported
 249                 by the device.
 250
 251 What:           /sys/block/<disk>/queue/zoned
 252 Date:           September 2016
 253 Contact:        Damien Le Moal <damien.lemoal@hgst.com>
 254 Description:
 255                 zoned indicates if the device is a zoned block device
 256                 and the zone model of the device if it is indeed zoned.
 257                 The possible values indicated by zoned are "none" for
 258                 regular block devices and "host-aware" or "host-managed"
 259                 for zoned block devices. The characteristics of
 260                 host-aware and host-managed zoned block devices are
 261                 described in the ZBC (Zoned Block Commands) and ZAC
 262                 (Zoned Device ATA Command Set) standards. These standards
 263                 also define the "drive-managed" zone model. However,
 264                 since drive-managed zoned block devices do not support
 265                 zone commands, they will be treated as regular block
 266                 devices and zoned will report "none".
 267
 268 What:           /sys/block/<disk>/queue/chunk_sectors
 269 Date:           September 2016
 270 Contact:        Hannes Reinecke <hare@suse.com>
 271 Description:
 272                 chunk_sectors has different meaning depending on the type
 273                 of the disk. For a RAID device (dm-raid), chunk_sectors
 274                 indicates the size in 512B sectors of the RAID volume
 275                 stripe segment. For a zoned block device, either
 276                 host-aware or host-managed, chunk_sectors indicates the
 277                 size of 512B sectors of the zones of the device, with
 278                 the eventual exception of the last zone of the device
 279                 which may be smaller.