1 What: /sys/block/<disk>/alignment_offset
3 Contact: Martin K. Petersen <martin.petersen@oracle.com>
5 Storage devices may report a physical block size that is
6 bigger than the logical block size (for instance a drive
7 with 4KB physical sectors exposing 512-byte logical
8 blocks to the operating system). This parameter
9 indicates how many bytes the beginning of the device is
10 offset from the disk's natural alignment.
13 What: /sys/block/<disk>/discard_alignment
15 Contact: Martin K. Petersen <martin.petersen@oracle.com>
17 Devices that support discard functionality may
18 internally allocate space in units that are bigger than
19 the exported logical block size. The discard_alignment
20 parameter indicates how many bytes the beginning of the
21 device is offset from the internal allocation unit's
25 What: /sys/block/<disk>/diskseq
27 Contact: Matteo Croce <mcroce@microsoft.com>
29 The /sys/block/<disk>/diskseq files reports the disk
30 sequence number, which is a monotonically increasing
31 number assigned to every drive.
32 Some devices, like the loop device, refresh such number
33 every time the backing file is changed.
34 The value type is 64 bit unsigned.
37 What: /sys/block/<disk>/inflight
39 Contact: Jens Axboe <axboe@kernel.dk>, Nikanth Karthikesan <knikanth@suse.de>
41 Reports the number of I/O requests currently in progress
42 (pending / in flight) in a device driver. This can be less
43 than the number of requests queued in the block device queue.
44 The report contains 2 fields: one for read requests
45 and one for write requests.
46 The value type is unsigned int.
47 Cf. Documentation/block/stat.rst which contains a single value for
49 This is related to /sys/block/<disk>/queue/nr_requests
50 and for SCSI device also its queue_depth.
53 What: /sys/block/<disk>/integrity/device_is_integrity_capable
55 Contact: Martin K. Petersen <martin.petersen@oracle.com>
57 Indicates whether a storage device is capable of storing
58 integrity metadata. Set if the device is T10 PI-capable.
61 What: /sys/block/<disk>/integrity/format
63 Contact: Martin K. Petersen <martin.petersen@oracle.com>
65 Metadata format for integrity capable block device.
66 E.g. T10-DIF-TYPE1-CRC.
69 What: /sys/block/<disk>/integrity/protection_interval_bytes
71 Contact: Martin K. Petersen <martin.petersen@oracle.com>
73 Describes the number of data bytes which are protected
74 by one integrity tuple. Typically the device's logical
78 What: /sys/block/<disk>/integrity/read_verify
80 Contact: Martin K. Petersen <martin.petersen@oracle.com>
82 Indicates whether the block layer should verify the
83 integrity of read requests serviced by devices that
84 support sending integrity metadata.
87 What: /sys/block/<disk>/integrity/tag_size
89 Contact: Martin K. Petersen <martin.petersen@oracle.com>
91 Number of bytes of integrity tag space available per
95 What: /sys/block/<disk>/integrity/write_generate
97 Contact: Martin K. Petersen <martin.petersen@oracle.com>
99 Indicates whether the block layer should automatically
100 generate checksums for write requests bound for
101 devices that support receiving integrity metadata.
104 What: /sys/block/<disk>/<partition>/alignment_offset
106 Contact: Martin K. Petersen <martin.petersen@oracle.com>
108 Storage devices may report a physical block size that is
109 bigger than the logical block size (for instance a drive
110 with 4KB physical sectors exposing 512-byte logical
111 blocks to the operating system). This parameter
112 indicates how many bytes the beginning of the partition
113 is offset from the disk's natural alignment.
116 What: /sys/block/<disk>/<partition>/discard_alignment
118 Contact: Martin K. Petersen <martin.petersen@oracle.com>
120 Devices that support discard functionality may
121 internally allocate space in units that are bigger than
122 the exported logical block size. The discard_alignment
123 parameter indicates how many bytes the beginning of the
124 partition is offset from the internal allocation unit's
128 What: /sys/block/<disk>/<partition>/stat
130 Contact: Jerome Marchand <jmarchan@redhat.com>
132 The /sys/block/<disk>/<partition>/stat files display the
133 I/O statistics of partition <partition>. The format is the
134 same as the format of /sys/block/<disk>/stat.
137 What: /sys/block/<disk>/queue/add_random
139 Contact: linux-block@vger.kernel.org
141 [RW] This file allows to turn off the disk entropy contribution.
142 Default value of this file is '1'(on).
145 What: /sys/block/<disk>/queue/chunk_sectors
147 Contact: Hannes Reinecke <hare@suse.com>
149 [RO] chunk_sectors has different meaning depending on the type
150 of the disk. For a RAID device (dm-raid), chunk_sectors
151 indicates the size in 512B sectors of the RAID volume stripe
152 segment. For a zoned block device, either host-aware or
153 host-managed, chunk_sectors indicates the size in 512B sectors
154 of the zones of the device, with the eventual exception of the
155 last zone of the device which may be smaller.
158 What: /sys/block/<disk>/queue/crypto/
160 Contact: linux-block@vger.kernel.org
162 The presence of this subdirectory of /sys/block/<disk>/queue/
163 indicates that the device supports inline encryption. This
164 subdirectory contains files which describe the inline encryption
165 capabilities of the device. For more information about inline
166 encryption, refer to Documentation/block/inline-encryption.rst.
169 What: /sys/block/<disk>/queue/crypto/max_dun_bits
171 Contact: linux-block@vger.kernel.org
173 [RO] This file shows the maximum length, in bits, of data unit
174 numbers accepted by the device in inline encryption requests.
177 What: /sys/block/<disk>/queue/crypto/modes/<mode>
179 Contact: linux-block@vger.kernel.org
181 [RO] For each crypto mode (i.e., encryption/decryption
182 algorithm) the device supports with inline encryption, a file
183 will exist at this location. It will contain a hexadecimal
184 number that is a bitmask of the supported data unit sizes, in
185 bytes, for that crypto mode.
187 Currently, the crypto modes that may be supported are:
193 For example, if a device supports AES-256-XTS inline encryption
194 with data unit sizes of 512 and 4096 bytes, the file
195 /sys/block/<disk>/queue/crypto/modes/AES-256-XTS will exist and
196 will contain "0x1200".
199 What: /sys/block/<disk>/queue/crypto/num_keyslots
201 Contact: linux-block@vger.kernel.org
203 [RO] This file shows the number of keyslots the device has for
204 use with inline encryption.
207 What: /sys/block/<disk>/queue/dax
209 Contact: linux-block@vger.kernel.org
211 [RO] This file indicates whether the device supports Direct
212 Access (DAX), used by CPU-addressable storage to bypass the
213 pagecache. It shows '1' if true, '0' if not.
216 What: /sys/block/<disk>/queue/discard_granularity
218 Contact: Martin K. Petersen <martin.petersen@oracle.com>
220 [RO] Devices that support discard functionality may internally
221 allocate space using units that are bigger than the logical
222 block size. The discard_granularity parameter indicates the size
223 of the internal allocation unit in bytes if reported by the
224 device. Otherwise the discard_granularity will be set to match
225 the device's physical block size. A discard_granularity of 0
226 means that the device does not support discard functionality.
229 What: /sys/block/<disk>/queue/discard_max_bytes
231 Contact: Martin K. Petersen <martin.petersen@oracle.com>
233 [RW] While discard_max_hw_bytes is the hardware limit for the
234 device, this setting is the software limit. Some devices exhibit
235 large latencies when large discards are issued, setting this
236 value lower will make Linux issue smaller discards and
237 potentially help reduce latencies induced by large discard
241 What: /sys/block/<disk>/queue/discard_max_hw_bytes
243 Contact: linux-block@vger.kernel.org
245 [RO] Devices that support discard functionality may have
246 internal limits on the number of bytes that can be trimmed or
247 unmapped in a single operation. The `discard_max_hw_bytes`
248 parameter is set by the device driver to the maximum number of
249 bytes that can be discarded in a single operation. Discard
250 requests issued to the device must not exceed this limit. A
251 `discard_max_hw_bytes` value of 0 means that the device does not
252 support discard functionality.
255 What: /sys/block/<disk>/queue/discard_zeroes_data
257 Contact: Martin K. Petersen <martin.petersen@oracle.com>
259 [RO] Will always return 0. Don't rely on any specific behavior
260 for discards, and don't read this file.
263 What: /sys/block/<disk>/queue/fua
265 Contact: linux-block@vger.kernel.org
267 [RO] Whether or not the block driver supports the FUA flag for
268 write requests. FUA stands for Force Unit Access. If the FUA
269 flag is set that means that write requests must bypass the
270 volatile cache of the storage device.
273 What: /sys/block/<disk>/queue/hw_sector_size
275 Contact: linux-block@vger.kernel.org
277 [RO] This is the hardware sector size of the device, in bytes.
280 What: /sys/block/<disk>/queue/independent_access_ranges/
282 Contact: linux-block@vger.kernel.org
284 [RO] The presence of this sub-directory of the
285 /sys/block/xxx/queue/ directory indicates that the device is
286 capable of executing requests targeting different sector ranges
287 in parallel. For instance, single LUN multi-actuator hard-disks
288 will have an independent_access_ranges directory if the device
289 correctly advertizes the sector ranges of its actuators.
291 The independent_access_ranges directory contains one directory
292 per access range, with each range described using the sector
293 (RO) attribute file to indicate the first sector of the range
294 and the nr_sectors (RO) attribute file to indicate the total
295 number of sectors in the range starting from the first sector of
296 the range. For example, a dual-actuator hard-disk will have the
297 following independent_access_ranges entries.::
299 $ tree /sys/block/<disk>/queue/independent_access_ranges/
300 /sys/block/<disk>/queue/independent_access_ranges/
308 The sector and nr_sectors attributes use 512B sector unit,
309 regardless of the actual block size of the device. Independent
310 access ranges do not overlap and include all sectors within the
311 device capacity. The access ranges are numbered in increasing
312 order of the range start sector, that is, the sector attribute
313 of range 0 always has the value 0.
316 What: /sys/block/<disk>/queue/io_poll
318 Contact: linux-block@vger.kernel.org
320 [RW] When read, this file shows whether polling is enabled (1)
321 or disabled (0). Writing '0' to this file will disable polling
322 for this device. Writing any non-zero value will enable this
326 What: /sys/block/<disk>/queue/io_poll_delay
328 Contact: linux-block@vger.kernel.org
330 [RW] If polling is enabled, this controls what kind of polling
331 will be performed. It defaults to -1, which is classic polling.
332 In this mode, the CPU will repeatedly ask for completions
333 without giving up any time. If set to 0, a hybrid polling mode
334 is used, where the kernel will attempt to make an educated guess
335 at when the IO will complete. Based on this guess, the kernel
336 will put the process issuing IO to sleep for an amount of time,
337 before entering a classic poll loop. This mode might be a little
338 slower than pure classic polling, but it will be more efficient.
339 If set to a value larger than 0, the kernel will put the process
340 issuing IO to sleep for this amount of microseconds before
341 entering classic polling.
344 What: /sys/block/<disk>/queue/io_timeout
346 Contact: Weiping Zhang <zhangweiping@didiglobal.com>
348 [RW] io_timeout is the request timeout in milliseconds. If a
349 request does not complete in this time then the block driver
350 timeout handler is invoked. That timeout handler can decide to
351 retry the request, to fail it or to start a device recovery
355 What: /sys/block/<disk>/queue/iostats
357 Contact: linux-block@vger.kernel.org
359 [RW] This file is used to control (on/off) the iostats
360 accounting of the disk.
363 What: /sys/block/<disk>/queue/logical_block_size
365 Contact: Martin K. Petersen <martin.petersen@oracle.com>
367 [RO] This is the smallest unit the storage device can address.
368 It is typically 512 bytes.
371 What: /sys/block/<disk>/queue/max_active_zones
373 Contact: Niklas Cassel <niklas.cassel@wdc.com>
375 [RO] For zoned block devices (zoned attribute indicating
376 "host-managed" or "host-aware"), the sum of zones belonging to
377 any of the zone states: EXPLICIT OPEN, IMPLICIT OPEN or CLOSED,
378 is limited by this value. If this value is 0, there is no limit.
380 If the host attempts to exceed this limit, the driver should
381 report this error with BLK_STS_ZONE_ACTIVE_RESOURCE, which user
382 space may see as the EOVERFLOW errno.
385 What: /sys/block/<disk>/queue/max_discard_segments
387 Contact: linux-block@vger.kernel.org
389 [RO] The maximum number of DMA scatter/gather entries in a
393 What: /sys/block/<disk>/queue/max_hw_sectors_kb
395 Contact: linux-block@vger.kernel.org
397 [RO] This is the maximum number of kilobytes supported in a
398 single data transfer.
401 What: /sys/block/<disk>/queue/max_integrity_segments
403 Contact: linux-block@vger.kernel.org
405 [RO] Maximum number of elements in a DMA scatter/gather list
406 with integrity data that will be submitted by the block layer
407 core to the associated block driver.
410 What: /sys/block/<disk>/queue/max_open_zones
412 Contact: Niklas Cassel <niklas.cassel@wdc.com>
414 [RO] For zoned block devices (zoned attribute indicating
415 "host-managed" or "host-aware"), the sum of zones belonging to
416 any of the zone states: EXPLICIT OPEN or IMPLICIT OPEN, is
417 limited by this value. If this value is 0, there is no limit.
420 What: /sys/block/<disk>/queue/max_sectors_kb
422 Contact: linux-block@vger.kernel.org
424 [RW] This is the maximum number of kilobytes that the block
425 layer will allow for a filesystem request. Must be smaller than
426 or equal to the maximum size allowed by the hardware.
429 What: /sys/block/<disk>/queue/max_segment_size
431 Contact: linux-block@vger.kernel.org
433 [RO] Maximum size in bytes of a single element in a DMA
437 What: /sys/block/<disk>/queue/max_segments
439 Contact: linux-block@vger.kernel.org
441 [RO] Maximum number of elements in a DMA scatter/gather list
442 that is submitted to the associated block driver.
445 What: /sys/block/<disk>/queue/minimum_io_size
447 Contact: Martin K. Petersen <martin.petersen@oracle.com>
449 [RO] Storage devices may report a granularity or preferred
450 minimum I/O size which is the smallest request the device can
451 perform without incurring a performance penalty. For disk
452 drives this is often the physical block size. For RAID arrays
453 it is often the stripe chunk size. A properly aligned multiple
454 of minimum_io_size is the preferred request size for workloads
455 where a high number of I/O operations is desired.
458 What: /sys/block/<disk>/queue/nomerges
460 Contact: linux-block@vger.kernel.org
462 [RW] Standard I/O elevator operations include attempts to merge
463 contiguous I/Os. For known random I/O loads these attempts will
464 always fail and result in extra cycles being spent in the
465 kernel. This allows one to turn off this behavior on one of two
466 ways: When set to 1, complex merge checks are disabled, but the
467 simple one-shot merges with the previous I/O request are
468 enabled. When set to 2, all merge tries are disabled. The
469 default value is 0 - which enables all types of merge tries.
472 What: /sys/block/<disk>/queue/nr_requests
474 Contact: linux-block@vger.kernel.org
476 [RW] This controls how many requests may be allocated in the
477 block layer for read or write requests. Note that the total
478 allocated number may be twice this amount, since it applies only
479 to reads or writes (not the accumulated sum).
481 To avoid priority inversion through request starvation, a
482 request queue maintains a separate request pool per each cgroup
483 when CONFIG_BLK_CGROUP is enabled, and this parameter applies to
484 each such per-block-cgroup request pool. IOW, if there are N
485 block cgroups, each request queue may have up to N request
486 pools, each independently regulated by nr_requests.
489 What: /sys/block/<disk>/queue/nr_zones
491 Contact: Damien Le Moal <damien.lemoal@wdc.com>
493 [RO] nr_zones indicates the total number of zones of a zoned
494 block device ("host-aware" or "host-managed" zone model). For
495 regular block devices, the value is always 0.
498 What: /sys/block/<disk>/queue/optimal_io_size
500 Contact: Martin K. Petersen <martin.petersen@oracle.com>
502 [RO] Storage devices may report an optimal I/O size, which is
503 the device's preferred unit for sustained I/O. This is rarely
504 reported for disk drives. For RAID arrays it is usually the
505 stripe width or the internal track size. A properly aligned
506 multiple of optimal_io_size is the preferred request size for
507 workloads where sustained throughput is desired. If no optimal
508 I/O size is reported this file contains 0.
511 What: /sys/block/<disk>/queue/physical_block_size
513 Contact: Martin K. Petersen <martin.petersen@oracle.com>
515 [RO] This is the smallest unit a physical storage device can
516 write atomically. It is usually the same as the logical block
517 size but may be bigger. One example is SATA drives with 4KB
518 sectors that expose a 512-byte logical block size to the
519 operating system. For stacked block devices the
520 physical_block_size variable contains the maximum
521 physical_block_size of the component devices.
524 What: /sys/block/<disk>/queue/read_ahead_kb
526 Contact: linux-block@vger.kernel.org
528 [RW] Maximum number of kilobytes to read-ahead for filesystems
529 on this block device.
532 What: /sys/block/<disk>/queue/rotational
534 Contact: linux-block@vger.kernel.org
536 [RW] This file is used to stat if the device is of rotational
537 type or non-rotational type.
540 What: /sys/block/<disk>/queue/rq_affinity
542 Contact: linux-block@vger.kernel.org
544 [RW] If this option is '1', the block layer will migrate request
545 completions to the cpu "group" that originally submitted the
546 request. For some workloads this provides a significant
547 reduction in CPU cycles due to caching effects.
549 For storage configurations that need to maximize distribution of
550 completion processing setting this option to '2' forces the
551 completion to run on the requesting cpu (bypassing the "group"
555 What: /sys/block/<disk>/queue/scheduler
557 Contact: linux-block@vger.kernel.org
559 [RW] When read, this file will display the current and available
560 IO schedulers for this block device. The currently active IO
561 scheduler will be enclosed in [] brackets. Writing an IO
562 scheduler name to this file will switch control of this block
563 device to that new IO scheduler. Note that writing an IO
564 scheduler name to this file will attempt to load that IO
565 scheduler module, if it isn't already present in the system.
568 What: /sys/block/<disk>/queue/stable_writes
570 Contact: linux-block@vger.kernel.org
572 [RW] This file will contain '1' if memory must not be modified
573 while it is being used in a write request to this device. When
574 this is the case and the kernel is performing writeback of a
575 page, the kernel will wait for writeback to complete before
576 allowing the page to be modified again, rather than allowing
577 immediate modification as is normally the case. This
578 restriction arises when the device accesses the memory multiple
579 times where the same data must be seen every time -- for
580 example, once to calculate a checksum and once to actually write
581 the data. If no such restriction exists, this file will contain
582 '0'. This file is writable for testing purposes.
585 What: /sys/block/<disk>/queue/throttle_sample_time
587 Contact: linux-block@vger.kernel.org
589 [RW] This is the time window that blk-throttle samples data, in
590 millisecond. blk-throttle makes decision based on the
591 samplings. Lower time means cgroups have more smooth throughput,
592 but higher CPU overhead. This exists only when
593 CONFIG_BLK_DEV_THROTTLING_LOW is enabled.
596 What: /sys/block/<disk>/queue/virt_boundary_mask
598 Contact: linux-block@vger.kernel.org
600 [RO] This file shows the I/O segment memory alignment mask for
601 the block device. I/O requests to this device will be split
602 between segments wherever either the memory address of the end
603 of the previous segment or the memory address of the beginning
604 of the current segment is not aligned to virt_boundary_mask + 1
608 What: /sys/block/<disk>/queue/wbt_lat_usec
610 Contact: linux-block@vger.kernel.org
612 [RW] If the device is registered for writeback throttling, then
613 this file shows the target minimum read latency. If this latency
614 is exceeded in a given window of time (see wb_window_usec), then
615 the writeback throttling will start scaling back writes. Writing
616 a value of '0' to this file disables the feature. Writing a
617 value of '-1' to this file resets the value to the default
621 What: /sys/block/<disk>/queue/write_cache
623 Contact: linux-block@vger.kernel.org
625 [RW] When read, this file will display whether the device has
626 write back caching enabled or not. It will return "write back"
627 for the former case, and "write through" for the latter. Writing
628 to this file can change the kernels view of the device, but it
629 doesn't alter the device state. This means that it might not be
630 safe to toggle the setting from "write back" to "write through",
631 since that will also eliminate cache flushes issued by the
635 What: /sys/block/<disk>/queue/write_same_max_bytes
637 Contact: Martin K. Petersen <martin.petersen@oracle.com>
639 [RO] Some devices support a write same operation in which a
640 single data block can be written to a range of several
641 contiguous blocks on storage. This can be used to wipe areas on
642 disk or to initialize drives in a RAID configuration.
643 write_same_max_bytes indicates how many bytes can be written in
644 a single write same command. If write_same_max_bytes is 0, write
645 same is not supported by the device.
648 What: /sys/block/<disk>/queue/write_zeroes_max_bytes
650 Contact: Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
652 [RO] Devices that support write zeroes operation in which a
653 single request can be issued to zero out the range of contiguous
654 blocks on storage without having any payload in the request.
655 This can be used to optimize writing zeroes to the devices.
656 write_zeroes_max_bytes indicates how many bytes can be written
657 in a single write zeroes command. If write_zeroes_max_bytes is
658 0, write zeroes is not supported by the device.
661 What: /sys/block/<disk>/queue/zone_append_max_bytes
663 Contact: linux-block@vger.kernel.org
665 [RO] This is the maximum number of bytes that can be written to
666 a sequential zone of a zoned block device using a zone append
667 write operation (REQ_OP_ZONE_APPEND). This value is always 0 for
668 regular block devices.
671 What: /sys/block/<disk>/queue/zone_write_granularity
673 Contact: linux-block@vger.kernel.org
675 [RO] This indicates the alignment constraint, in bytes, for
676 write operations in sequential zones of zoned block devices
677 (devices with a zoned attributed that reports "host-managed" or
678 "host-aware"). This value is always 0 for regular block devices.
681 What: /sys/block/<disk>/queue/zoned
683 Contact: Damien Le Moal <damien.lemoal@wdc.com>
685 [RO] zoned indicates if the device is a zoned block device and
686 the zone model of the device if it is indeed zoned. The
687 possible values indicated by zoned are "none" for regular block
688 devices and "host-aware" or "host-managed" for zoned block
689 devices. The characteristics of host-aware and host-managed
690 zoned block devices are described in the ZBC (Zoned Block
691 Commands) and ZAC (Zoned Device ATA Command Set) standards.
692 These standards also define the "drive-managed" zone model.
693 However, since drive-managed zoned block devices do not support
694 zone commands, they will be treated as regular block devices and
695 zoned will report "none".
698 What: /sys/block/<disk>/stat
700 Contact: Jerome Marchand <jmarchan@redhat.com>
702 The /sys/block/<disk>/stat files displays the I/O
703 statistics of disk <disk>. They contain 11 fields:
705 == ==============================================
706 1 reads completed successfully
709 4 time spent reading (ms)
713 8 time spent writing (ms)
714 9 I/Os currently in progress
715 10 time spent doing I/Os (ms)
716 11 weighted time spent doing I/Os (ms)
717 12 discards completed
720 15 time spent discarding (ms)
721 16 flush requests completed
722 17 time spent flushing (ms)
723 == ==============================================
725 For more details refer Documentation/admin-guide/iostats.rst