2 .\" Copyright (c) 2013 by Turbo Fredriksson <turbo@bayour.com>. All rights reserved.
3 .\" Copyright (c) 2017 Datto Inc.
4 .\" The contents of this file are subject to the terms of the Common Development
5 .\" and Distribution License (the "License"). You may not use this file except
6 .\" in compliance with the License. You can obtain a copy of the license at
7 .\" usr/src/OPENSOLARIS.LICENSE or http://www.opensolaris.org/os/licensing.
9 .\" See the License for the specific language governing permissions and
10 .\" limitations under the License. When distributing Covered Code, include this
11 .\" CDDL HEADER in each file and include the License file at
12 .\" usr/src/OPENSOLARIS.LICENSE. If applicable, add the following below this
13 .\" CDDL HEADER, with the fields enclosed by brackets "[]" replaced with your
14 .\" own identifying information:
15 .\" Portions Copyright [yyyy] [name of copyright owner]
16 .TH ZFS-MODULE-PARAMETERS 5 "Oct 28, 2017"
18 zfs\-module\-parameters \- ZFS module parameters
22 Description of the different parameters to the ZFS module.
24 .SS "Module parameters"
31 \fBignore_hole_birth\fR (int)
34 When set, the hole_birth optimization will not be used, and all holes will
35 always be sent on zfs send. Useful if you suspect your datasets are affected
36 by a bug in hole_birth.
38 Use \fB1\fR for on (default) and \fB0\fR for off.
44 \fBl2arc_feed_again\fR (int)
47 Turbo L2ARC warm-up. When the L2ARC is cold the fill interval will be set as
50 Use \fB1\fR for yes (default) and \fB0\fR to disable.
56 \fBl2arc_feed_min_ms\fR (ulong)
59 Min feed interval in milliseconds. Requires \fBl2arc_feed_again=1\fR and only
60 applicable in related situations.
62 Default value: \fB200\fR.
68 \fBl2arc_feed_secs\fR (ulong)
71 Seconds between L2ARC writing
73 Default value: \fB1\fR.
79 \fBl2arc_headroom\fR (ulong)
82 How far through the ARC lists to search for L2ARC cacheable content, expressed
83 as a multiplier of \fBl2arc_write_max\fR
85 Default value: \fB2\fR.
91 \fBl2arc_headroom_boost\fR (ulong)
94 Scales \fBl2arc_headroom\fR by this percentage when L2ARC contents are being
95 successfully compressed before writing. A value of 100 disables this feature.
97 Default value: \fB200\fR%.
103 \fBl2arc_noprefetch\fR (int)
106 Do not write buffers to L2ARC if they were prefetched but not used by
109 Use \fB1\fR for yes (default) and \fB0\fR to disable.
115 \fBl2arc_norw\fR (int)
118 No reads during writes
120 Use \fB1\fR for yes and \fB0\fR for no (default).
126 \fBl2arc_write_boost\fR (ulong)
129 Cold L2ARC devices will have \fBl2arc_write_max\fR increased by this amount
130 while they remain cold.
132 Default value: \fB8,388,608\fR.
138 \fBl2arc_write_max\fR (ulong)
141 Max write bytes per interval
143 Default value: \fB8,388,608\fR.
149 \fBmetaslab_aliquot\fR (ulong)
152 Metaslab granularity, in bytes. This is roughly similar to what would be
153 referred to as the "stripe size" in traditional RAID arrays. In normal
154 operation, ZFS will try to write this amount of data to a top-level vdev
155 before moving on to the next one.
157 Default value: \fB524,288\fR.
163 \fBmetaslab_bias_enabled\fR (int)
166 Enable metaslab group biasing based on its vdev's over- or under-utilization
167 relative to the pool.
169 Use \fB1\fR for yes (default) and \fB0\fR for no.
175 \fBzfs_metaslab_segment_weight_enabled\fR (int)
178 Enable/disable segment-based metaslab selection.
180 Use \fB1\fR for yes (default) and \fB0\fR for no.
186 \fBzfs_metaslab_switch_threshold\fR (int)
189 When using segment-based metaslab selection, continue allocating
190 from the active metaslab until \fBzfs_metaslab_switch_threshold\fR
191 worth of buckets have been exhausted.
193 Default value: \fB2\fR.
199 \fBmetaslab_debug_load\fR (int)
202 Load all metaslabs during pool import.
204 Use \fB1\fR for yes and \fB0\fR for no (default).
210 \fBmetaslab_debug_unload\fR (int)
213 Prevent metaslabs from being unloaded.
215 Use \fB1\fR for yes and \fB0\fR for no (default).
221 \fBmetaslab_fragmentation_factor_enabled\fR (int)
224 Enable use of the fragmentation metric in computing metaslab weights.
226 Use \fB1\fR for yes (default) and \fB0\fR for no.
232 \fBmetaslabs_per_vdev\fR (int)
235 When a vdev is added, it will be divided into approximately (but no more than) this number of metaslabs.
237 Default value: \fB200\fR.
243 \fBmetaslab_preload_enabled\fR (int)
246 Enable metaslab group preloading.
248 Use \fB1\fR for yes (default) and \fB0\fR for no.
254 \fBmetaslab_lba_weighting_enabled\fR (int)
257 Give more weight to metaslabs with lower LBAs, assuming they have
258 greater bandwidth as is typically the case on a modern constant
259 angular velocity disk drive.
261 Use \fB1\fR for yes (default) and \fB0\fR for no.
267 \fBspa_config_path\fR (charp)
272 Default value: \fB/etc/zfs/zpool.cache\fR.
278 \fBspa_asize_inflation\fR (int)
281 Multiplication factor used to estimate actual disk consumption from the
282 size of data being written. The default value is a worst case estimate,
283 but lower values may be valid for a given pool depending on its
284 configuration. Pool administrators who understand the factors involved
285 may wish to specify a more realistic inflation factor, particularly if
286 they operate close to quota or capacity limits.
288 Default value: \fB24\fR.
294 \fBspa_load_verify_data\fR (int)
297 Whether to traverse data blocks during an "extreme rewind" (\fB-X\fR)
298 import. Use 0 to disable and 1 to enable.
300 An extreme rewind import normally performs a full traversal of all
301 blocks in the pool for verification. If this parameter is set to 0,
302 the traversal skips non-metadata blocks. It can be toggled once the
303 import has started to stop or start the traversal of non-metadata blocks.
305 Default value: \fB1\fR.
311 \fBspa_load_verify_metadata\fR (int)
314 Whether to traverse blocks during an "extreme rewind" (\fB-X\fR)
315 pool import. Use 0 to disable and 1 to enable.
317 An extreme rewind import normally performs a full traversal of all
318 blocks in the pool for verification. If this parameter is set to 0,
319 the traversal is not performed. It can be toggled once the import has
320 started to stop or start the traversal.
322 Default value: \fB1\fR.
328 \fBspa_load_verify_maxinflight\fR (int)
331 Maximum concurrent I/Os during the traversal performed during an "extreme
332 rewind" (\fB-X\fR) pool import.
334 Default value: \fB10000\fR.
340 \fBspa_slop_shift\fR (int)
343 Normally, we don't allow the last 3.2% (1/(2^spa_slop_shift)) of space
344 in the pool to be consumed. This ensures that we don't run the pool
345 completely out of space, due to unaccounted changes (e.g. to the MOS).
346 It also limits the worst-case time to allocate space. If we have
347 less than this amount of free space, most ZPL operations (e.g. write,
348 create) will return ENOSPC.
350 Default value: \fB5\fR.
356 \fBzfetch_array_rd_sz\fR (ulong)
359 If prefetching is enabled, disable prefetching for reads larger than this size.
361 Default value: \fB1,048,576\fR.
367 \fBzfetch_max_distance\fR (uint)
370 Max bytes to prefetch per stream (default 8MB).
372 Default value: \fB8,388,608\fR.
378 \fBzfetch_max_streams\fR (uint)
381 Max number of streams per zfetch (prefetch streams per file).
383 Default value: \fB8\fR.
389 \fBzfetch_min_sec_reap\fR (uint)
392 Min time before an active prefetch stream can be reclaimed
394 Default value: \fB2\fR.
400 \fBzfs_arc_dnode_limit\fR (ulong)
403 When the number of bytes consumed by dnodes in the ARC exceeds this number of
404 bytes, try to unpin some of it in response to demand for non-metadata. This
405 value acts as a ceiling to the amount of dnode metadata, and defaults to 0 which
406 indicates that a percent which is based on \fBzfs_arc_dnode_limit_percent\fR of
407 the ARC meta buffers that may be used for dnodes.
409 See also \fBzfs_arc_meta_prune\fR which serves a similar purpose but is used
410 when the amount of metadata in the ARC exceeds \fBzfs_arc_meta_limit\fR rather
411 than in response to overall demand for non-metadata.
414 Default value: \fB0\fR.
420 \fBzfs_arc_dnode_limit_percent\fR (ulong)
423 Percentage that can be consumed by dnodes of ARC meta buffers.
425 See also \fBzfs_arc_dnode_limit\fR which serves a similar purpose but has a
426 higher priority if set to nonzero value.
428 Default value: \fB10\fR%.
434 \fBzfs_arc_dnode_reduce_percent\fR (ulong)
437 Percentage of ARC dnodes to try to scan in response to demand for non-metadata
438 when the number of bytes consumed by dnodes exceeds \fBzfs_arc_dnode_limit\fR.
441 Default value: \fB10\fR% of the number of dnodes in the ARC.
447 \fBzfs_arc_average_blocksize\fR (int)
450 The ARC's buffer hash table is sized based on the assumption of an average
451 block size of \fBzfs_arc_average_blocksize\fR (default 8K). This works out
452 to roughly 1MB of hash table per 1GB of physical memory with 8-byte pointers.
453 For configurations with a known larger average block size this value can be
454 increased to reduce the memory footprint.
457 Default value: \fB8192\fR.
463 \fBzfs_arc_evict_batch_limit\fR (int)
466 Number ARC headers to evict per sub-list before proceeding to another sub-list.
467 This batch-style operation prevents entire sub-lists from being evicted at once
468 but comes at a cost of additional unlocking and locking.
470 Default value: \fB10\fR.
476 \fBzfs_arc_grow_retry\fR (int)
479 If set to a non zero value, it will replace the arc_grow_retry value with this value.
480 The arc_grow_retry value (default 5) is the number of seconds the ARC will wait before
481 trying to resume growth after a memory pressure event.
483 Default value: \fB0\fR.
489 \fBzfs_arc_lotsfree_percent\fR (int)
492 Throttle I/O when free system memory drops below this percentage of total
493 system memory. Setting this value to 0 will disable the throttle.
495 Default value: \fB10\fR%.
501 \fBzfs_arc_max\fR (ulong)
504 Max arc size of ARC in bytes. If set to 0 then it will consume 1/2 of system
505 RAM. This value must be at least 67108864 (64 megabytes).
507 This value can be changed dynamically with some caveats. It cannot be set back
508 to 0 while running and reducing it below the current ARC size will not cause
509 the ARC to shrink without memory pressure to induce shrinking.
511 Default value: \fB0\fR.
517 \fBzfs_arc_meta_adjust_restarts\fR (ulong)
520 The number of restart passes to make while scanning the ARC attempting
521 the free buffers in order to stay below the \fBzfs_arc_meta_limit\fR.
522 This value should not need to be tuned but is available to facilitate
523 performance analysis.
525 Default value: \fB4096\fR.
531 \fBzfs_arc_meta_limit\fR (ulong)
534 The maximum allowed size in bytes that meta data buffers are allowed to
535 consume in the ARC. When this limit is reached meta data buffers will
536 be reclaimed even if the overall arc_c_max has not been reached. This
537 value defaults to 0 which indicates that a percent which is based on
538 \fBzfs_arc_meta_limit_percent\fR of the ARC may be used for meta data.
540 This value my be changed dynamically except that it cannot be set back to 0
541 for a specific percent of the ARC; it must be set to an explicit value.
543 Default value: \fB0\fR.
549 \fBzfs_arc_meta_limit_percent\fR (ulong)
552 Percentage of ARC buffers that can be used for meta data.
554 See also \fBzfs_arc_meta_limit\fR which serves a similar purpose but has a
555 higher priority if set to nonzero value.
558 Default value: \fB75\fR%.
564 \fBzfs_arc_meta_min\fR (ulong)
567 The minimum allowed size in bytes that meta data buffers may consume in
568 the ARC. This value defaults to 0 which disables a floor on the amount
569 of the ARC devoted meta data.
571 Default value: \fB0\fR.
577 \fBzfs_arc_meta_prune\fR (int)
580 The number of dentries and inodes to be scanned looking for entries
581 which can be dropped. This may be required when the ARC reaches the
582 \fBzfs_arc_meta_limit\fR because dentries and inodes can pin buffers
583 in the ARC. Increasing this value will cause to dentry and inode caches
584 to be pruned more aggressively. Setting this value to 0 will disable
585 pruning the inode and dentry caches.
587 Default value: \fB10,000\fR.
593 \fBzfs_arc_meta_strategy\fR (int)
596 Define the strategy for ARC meta data buffer eviction (meta reclaim strategy).
597 A value of 0 (META_ONLY) will evict only the ARC meta data buffers.
598 A value of 1 (BALANCED) indicates that additional data buffers may be evicted if
599 that is required to in order to evict the required number of meta data buffers.
601 Default value: \fB1\fR.
607 \fBzfs_arc_min\fR (ulong)
610 Min arc size of ARC in bytes. If set to 0 then arc_c_min will default to
611 consuming the larger of 32M or 1/32 of total system memory.
613 Default value: \fB0\fR.
619 \fBzfs_arc_min_prefetch_ms\fR (int)
622 Minimum time prefetched blocks are locked in the ARC, specified in ms.
623 A value of \fB0\fR will default to 1 second.
625 Default value: \fB0\fR.
631 \fBzfs_arc_min_prescient_prefetch_ms\fR (int)
634 Minimum time "prescient prefetched" blocks are locked in the ARC, specified
635 in ms. These blocks are meant to be prefetched fairly aggresively ahead of
636 the code that may use them. A value of \fB0\fR will default to 6 seconds.
638 Default value: \fB0\fR.
644 \fBzfs_multilist_num_sublists\fR (int)
647 To allow more fine-grained locking, each ARC state contains a series
648 of lists for both data and meta data objects. Locking is performed at
649 the level of these "sub-lists". This parameters controls the number of
650 sub-lists per ARC state, and also applies to other uses of the
651 multilist data structure.
653 Default value: \fB4\fR or the number of online CPUs, whichever is greater
659 \fBzfs_arc_overflow_shift\fR (int)
662 The ARC size is considered to be overflowing if it exceeds the current
663 ARC target size (arc_c) by a threshold determined by this parameter.
664 The threshold is calculated as a fraction of arc_c using the formula
665 "arc_c >> \fBzfs_arc_overflow_shift\fR".
667 The default value of 8 causes the ARC to be considered to be overflowing
668 if it exceeds the target size by 1/256th (0.3%) of the target size.
670 When the ARC is overflowing, new buffer allocations are stalled until
671 the reclaim thread catches up and the overflow condition no longer exists.
673 Default value: \fB8\fR.
680 \fBzfs_arc_p_min_shift\fR (int)
683 If set to a non zero value, this will update arc_p_min_shift (default 4)
685 arc_p_min_shift is used to shift of arc_c for calculating both min and max
688 Default value: \fB0\fR.
694 \fBzfs_arc_p_aggressive_disable\fR (int)
697 Disable aggressive arc_p growth
699 Use \fB1\fR for yes (default) and \fB0\fR to disable.
705 \fBzfs_arc_p_dampener_disable\fR (int)
708 Disable arc_p adapt dampener
710 Use \fB1\fR for yes (default) and \fB0\fR to disable.
716 \fBzfs_arc_shrink_shift\fR (int)
719 If set to a non zero value, this will update arc_shrink_shift (default 7)
722 Default value: \fB0\fR.
728 \fBzfs_arc_pc_percent\fR (uint)
731 Percent of pagecache to reclaim arc to
733 This tunable allows ZFS arc to play more nicely with the kernel's LRU
734 pagecache. It can guarantee that the arc size won't collapse under scanning
735 pressure on the pagecache, yet still allows arc to be reclaimed down to
736 zfs_arc_min if necessary. This value is specified as percent of pagecache
737 size (as measured by NR_FILE_PAGES) where that percent may exceed 100. This
738 only operates during memory pressure/reclaim.
740 Default value: \fB0\fR% (disabled).
746 \fBzfs_arc_sys_free\fR (ulong)
749 The target number of bytes the ARC should leave as free memory on the system.
750 Defaults to the larger of 1/64 of physical memory or 512K. Setting this
751 option to a non-zero value will override the default.
753 Default value: \fB0\fR.
759 \fBzfs_autoimport_disable\fR (int)
762 Disable pool import at module load by ignoring the cache file (typically \fB/etc/zfs/zpool.cache\fR).
764 Use \fB1\fR for yes (default) and \fB0\fR for no.
770 \fBzfs_commit_timeout_pct\fR (int)
773 This controls the amount of time that a ZIL block (lwb) will remain "open"
774 when it isn't "full", and it has a thread waiting for it to be committed to
775 stable storage. The timeout is scaled based on a percentage of the last lwb
776 latency to avoid significantly impacting the latency of each individual
777 transaction record (itx).
779 Default value: \fB5\fR%.
785 \fBzfs_dbgmsg_enable\fR (int)
788 Internally ZFS keeps a small log to facilitate debugging. By default the log
789 is disabled, to enable it set this option to 1. The contents of the log can
790 be accessed by reading the /proc/spl/kstat/zfs/dbgmsg file. Writing 0 to
791 this proc file clears the log.
793 Default value: \fB0\fR.
799 \fBzfs_dbgmsg_maxsize\fR (int)
802 The maximum size in bytes of the internal ZFS debug log.
804 Default value: \fB4M\fR.
810 \fBzfs_dbuf_state_index\fR (int)
813 This feature is currently unused. It is normally used for controlling what
814 reporting is available under /proc/spl/kstat/zfs.
816 Default value: \fB0\fR.
822 \fBzfs_deadman_enabled\fR (int)
825 When a pool sync operation takes longer than \fBzfs_deadman_synctime_ms\fR
826 milliseconds, a "slow spa_sync" message is logged to the debug log
827 (see \fBzfs_dbgmsg_enable\fR). If \fBzfs_deadman_enabled\fR is set,
828 all pending IO operations are also checked and if any haven't completed
829 within \fBzfs_deadman_synctime_ms\fR milliseconds, a "SLOW IO" message
830 is logged to the debug log and a "delay" system event with the details of
831 the hung IO is posted.
833 Use \fB1\fR (default) to enable the slow IO check and \fB0\fR to disable.
839 \fBzfs_deadman_checktime_ms\fR (int)
842 Once a pool sync operation has taken longer than
843 \fBzfs_deadman_synctime_ms\fR milliseconds, continue to check for slow
844 operations every \fBzfs_deadman_checktime_ms\fR milliseconds.
846 Default value: \fB5,000\fR.
852 \fBzfs_deadman_synctime_ms\fR (ulong)
855 Interval in milliseconds after which the deadman is triggered and also
856 the interval after which an IO operation is considered to be "hung"
857 if \fBzfs_deadman_enabled\fR is set.
859 See \fBzfs_deadman_enabled\fR.
861 Default value: \fB1,000,000\fR.
867 \fBzfs_dedup_prefetch\fR (int)
870 Enable prefetching dedup-ed blks
872 Use \fB1\fR for yes and \fB0\fR to disable (default).
878 \fBzfs_delay_min_dirty_percent\fR (int)
881 Start to delay each transaction once there is this amount of dirty data,
882 expressed as a percentage of \fBzfs_dirty_data_max\fR.
883 This value should be >= zfs_vdev_async_write_active_max_dirty_percent.
884 See the section "ZFS TRANSACTION DELAY".
886 Default value: \fB60\fR%.
892 \fBzfs_delay_scale\fR (int)
895 This controls how quickly the transaction delay approaches infinity.
896 Larger values cause longer delays for a given amount of dirty data.
898 For the smoothest delay, this value should be about 1 billion divided
899 by the maximum number of operations per second. This will smoothly
900 handle between 10x and 1/10th this number.
902 See the section "ZFS TRANSACTION DELAY".
904 Note: \fBzfs_delay_scale\fR * \fBzfs_dirty_data_max\fR must be < 2^64.
906 Default value: \fB500,000\fR.
912 \fBzfs_delete_blocks\fR (ulong)
915 This is the used to define a large file for the purposes of delete. Files
916 containing more than \fBzfs_delete_blocks\fR will be deleted asynchronously
917 while smaller files are deleted synchronously. Decreasing this value will
918 reduce the time spent in an unlink(2) system call at the expense of a longer
919 delay before the freed space is available.
921 Default value: \fB20,480\fR.
927 \fBzfs_dirty_data_max\fR (int)
930 Determines the dirty space limit in bytes. Once this limit is exceeded, new
931 writes are halted until space frees up. This parameter takes precedence
932 over \fBzfs_dirty_data_max_percent\fR.
933 See the section "ZFS TRANSACTION DELAY".
935 Default value: \fB10\fR% of physical RAM, capped at \fBzfs_dirty_data_max_max\fR.
941 \fBzfs_dirty_data_max_max\fR (int)
944 Maximum allowable value of \fBzfs_dirty_data_max\fR, expressed in bytes.
945 This limit is only enforced at module load time, and will be ignored if
946 \fBzfs_dirty_data_max\fR is later changed. This parameter takes
947 precedence over \fBzfs_dirty_data_max_max_percent\fR. See the section
948 "ZFS TRANSACTION DELAY".
950 Default value: \fB25\fR% of physical RAM.
956 \fBzfs_dirty_data_max_max_percent\fR (int)
959 Maximum allowable value of \fBzfs_dirty_data_max\fR, expressed as a
960 percentage of physical RAM. This limit is only enforced at module load
961 time, and will be ignored if \fBzfs_dirty_data_max\fR is later changed.
962 The parameter \fBzfs_dirty_data_max_max\fR takes precedence over this
963 one. See the section "ZFS TRANSACTION DELAY".
965 Default value: \fB25\fR%.
971 \fBzfs_dirty_data_max_percent\fR (int)
974 Determines the dirty space limit, expressed as a percentage of all
975 memory. Once this limit is exceeded, new writes are halted until space frees
976 up. The parameter \fBzfs_dirty_data_max\fR takes precedence over this
977 one. See the section "ZFS TRANSACTION DELAY".
979 Default value: \fB10\fR%, subject to \fBzfs_dirty_data_max_max\fR.
985 \fBzfs_dirty_data_sync\fR (int)
988 Start syncing out a transaction group if there is at least this much dirty data.
990 Default value: \fB67,108,864\fR.
996 \fBzfs_fletcher_4_impl\fR (string)
999 Select a fletcher 4 implementation.
1001 Supported selectors are: \fBfastest\fR, \fBscalar\fR, \fBsse2\fR, \fBssse3\fR,
1002 \fBavx2\fR, \fBavx512f\fR, and \fBaarch64_neon\fR.
1003 All of the selectors except \fBfastest\fR and \fBscalar\fR require instruction
1004 set extensions to be available and will only appear if ZFS detects that they are
1005 present at runtime. If multiple implementations of fletcher 4 are available,
1006 the \fBfastest\fR will be chosen using a micro benchmark. Selecting \fBscalar\fR
1007 results in the original, CPU based calculation, being used. Selecting any option
1008 other than \fBfastest\fR and \fBscalar\fR results in vector instructions from
1009 the respective CPU instruction set being used.
1011 Default value: \fBfastest\fR.
1017 \fBzfs_free_bpobj_enabled\fR (int)
1020 Enable/disable the processing of the free_bpobj object.
1022 Default value: \fB1\fR.
1028 \fBzfs_free_max_blocks\fR (ulong)
1031 Maximum number of blocks freed in a single txg.
1033 Default value: \fB100,000\fR.
1039 \fBzfs_vdev_async_read_max_active\fR (int)
1042 Maximum asynchronous read I/Os active to each device.
1043 See the section "ZFS I/O SCHEDULER".
1045 Default value: \fB3\fR.
1051 \fBzfs_vdev_async_read_min_active\fR (int)
1054 Minimum asynchronous read I/Os active to each device.
1055 See the section "ZFS I/O SCHEDULER".
1057 Default value: \fB1\fR.
1063 \fBzfs_vdev_async_write_active_max_dirty_percent\fR (int)
1066 When the pool has more than
1067 \fBzfs_vdev_async_write_active_max_dirty_percent\fR dirty data, use
1068 \fBzfs_vdev_async_write_max_active\fR to limit active async writes. If
1069 the dirty data is between min and max, the active I/O limit is linearly
1070 interpolated. See the section "ZFS I/O SCHEDULER".
1072 Default value: \fB60\fR%.
1078 \fBzfs_vdev_async_write_active_min_dirty_percent\fR (int)
1081 When the pool has less than
1082 \fBzfs_vdev_async_write_active_min_dirty_percent\fR dirty data, use
1083 \fBzfs_vdev_async_write_min_active\fR to limit active async writes. If
1084 the dirty data is between min and max, the active I/O limit is linearly
1085 interpolated. See the section "ZFS I/O SCHEDULER".
1087 Default value: \fB30\fR%.
1093 \fBzfs_vdev_async_write_max_active\fR (int)
1096 Maximum asynchronous write I/Os active to each device.
1097 See the section "ZFS I/O SCHEDULER".
1099 Default value: \fB10\fR.
1105 \fBzfs_vdev_async_write_min_active\fR (int)
1108 Minimum asynchronous write I/Os active to each device.
1109 See the section "ZFS I/O SCHEDULER".
1111 Lower values are associated with better latency on rotational media but poorer
1112 resilver performance. The default value of 2 was chosen as a compromise. A
1113 value of 3 has been shown to improve resilver performance further at a cost of
1114 further increasing latency.
1116 Default value: \fB2\fR.
1122 \fBzfs_vdev_max_active\fR (int)
1125 The maximum number of I/Os active to each device. Ideally, this will be >=
1126 the sum of each queue's max_active. It must be at least the sum of each
1127 queue's min_active. See the section "ZFS I/O SCHEDULER".
1129 Default value: \fB1,000\fR.
1135 \fBzfs_vdev_scrub_max_active\fR (int)
1138 Maximum scrub I/Os active to each device.
1139 See the section "ZFS I/O SCHEDULER".
1141 Default value: \fB2\fR.
1147 \fBzfs_vdev_scrub_min_active\fR (int)
1150 Minimum scrub I/Os active to each device.
1151 See the section "ZFS I/O SCHEDULER".
1153 Default value: \fB1\fR.
1159 \fBzfs_vdev_sync_read_max_active\fR (int)
1162 Maximum synchronous read I/Os active to each device.
1163 See the section "ZFS I/O SCHEDULER".
1165 Default value: \fB10\fR.
1171 \fBzfs_vdev_sync_read_min_active\fR (int)
1174 Minimum synchronous read I/Os active to each device.
1175 See the section "ZFS I/O SCHEDULER".
1177 Default value: \fB10\fR.
1183 \fBzfs_vdev_sync_write_max_active\fR (int)
1186 Maximum synchronous write I/Os active to each device.
1187 See the section "ZFS I/O SCHEDULER".
1189 Default value: \fB10\fR.
1195 \fBzfs_vdev_sync_write_min_active\fR (int)
1198 Minimum synchronous write I/Os active to each device.
1199 See the section "ZFS I/O SCHEDULER".
1201 Default value: \fB10\fR.
1207 \fBzfs_vdev_queue_depth_pct\fR (int)
1210 Maximum number of queued allocations per top-level vdev expressed as
1211 a percentage of \fBzfs_vdev_async_write_max_active\fR which allows the
1212 system to detect devices that are more capable of handling allocations
1213 and to allocate more blocks to those devices. It allows for dynamic
1214 allocation distribution when devices are imbalanced as fuller devices
1215 will tend to be slower than empty devices.
1217 See also \fBzio_dva_throttle_enabled\fR.
1219 Default value: \fB1000\fR%.
1225 \fBzfs_disable_dup_eviction\fR (int)
1228 Disable duplicate buffer eviction
1230 Use \fB1\fR for yes and \fB0\fR for no (default).
1236 \fBzfs_expire_snapshot\fR (int)
1239 Seconds to expire .zfs/snapshot
1241 Default value: \fB300\fR.
1247 \fBzfs_admin_snapshot\fR (int)
1250 Allow the creation, removal, or renaming of entries in the .zfs/snapshot
1251 directory to cause the creation, destruction, or renaming of snapshots.
1252 When enabled this functionality works both locally and over NFS exports
1253 which have the 'no_root_squash' option set. This functionality is disabled
1256 Use \fB1\fR for yes and \fB0\fR for no (default).
1262 \fBzfs_flags\fR (int)
1265 Set additional debugging flags. The following flags may be bitwise-or'd
1277 Enable dprintf entries in the debug log.
1279 2 ZFS_DEBUG_DBUF_VERIFY *
1280 Enable extra dbuf verifications.
1282 4 ZFS_DEBUG_DNODE_VERIFY *
1283 Enable extra dnode verifications.
1285 8 ZFS_DEBUG_SNAPNAMES
1286 Enable snapshot name verification.
1289 Check for illegally modified ARC buffers.
1292 Enable spa_dbgmsg entries in the debug log.
1294 64 ZFS_DEBUG_ZIO_FREE
1295 Enable verification of block frees.
1297 128 ZFS_DEBUG_HISTOGRAM_VERIFY
1298 Enable extra spacemap histogram verifications.
1300 256 ZFS_DEBUG_METASLAB_VERIFY
1301 Verify space accounting on disk matches in-core range_trees.
1303 512 ZFS_DEBUG_SET_ERROR
1304 Enable SET_ERROR and dprintf entries in the debug log.
1307 * Requires debug build.
1309 Default value: \fB0\fR.
1315 \fBzfs_free_leak_on_eio\fR (int)
1318 If destroy encounters an EIO while reading metadata (e.g. indirect
1319 blocks), space referenced by the missing metadata can not be freed.
1320 Normally this causes the background destroy to become "stalled", as
1321 it is unable to make forward progress. While in this stalled state,
1322 all remaining space to free from the error-encountering filesystem is
1323 "temporarily leaked". Set this flag to cause it to ignore the EIO,
1324 permanently leak the space from indirect blocks that can not be read,
1325 and continue to free everything else that it can.
1327 The default, "stalling" behavior is useful if the storage partially
1328 fails (i.e. some but not all i/os fail), and then later recovers. In
1329 this case, we will be able to continue pool operations while it is
1330 partially failed, and when it recovers, we can continue to free the
1331 space, with no leaks. However, note that this case is actually
1334 Typically pools either (a) fail completely (but perhaps temporarily,
1335 e.g. a top-level vdev going offline), or (b) have localized,
1336 permanent errors (e.g. disk returns the wrong data due to bit flip or
1337 firmware bug). In case (a), this setting does not matter because the
1338 pool will be suspended and the sync thread will not be able to make
1339 forward progress regardless. In case (b), because the error is
1340 permanent, the best we can do is leak the minimum amount of space,
1341 which is what setting this flag will do. Therefore, it is reasonable
1342 for this flag to normally be set, but we chose the more conservative
1343 approach of not setting it, so that there is no possibility of
1344 leaking space in the "partial temporary" failure case.
1346 Default value: \fB0\fR.
1352 \fBzfs_free_min_time_ms\fR (int)
1355 During a \fBzfs destroy\fR operation using \fBfeature@async_destroy\fR a minimum
1356 of this much time will be spent working on freeing blocks per txg.
1358 Default value: \fB1,000\fR.
1364 \fBzfs_immediate_write_sz\fR (long)
1367 Largest data block to write to zil. Larger blocks will be treated as if the
1368 dataset being written to had the property setting \fBlogbias=throughput\fR.
1370 Default value: \fB32,768\fR.
1376 \fBzfs_max_recordsize\fR (int)
1379 We currently support block sizes from 512 bytes to 16MB. The benefits of
1380 larger blocks, and thus larger IO, need to be weighed against the cost of
1381 COWing a giant block to modify one byte. Additionally, very large blocks
1382 can have an impact on i/o latency, and also potentially on the memory
1383 allocator. Therefore, we do not allow the recordsize to be set larger than
1384 zfs_max_recordsize (default 1MB). Larger blocks can be created by changing
1385 this tunable, and pools with larger blocks can always be imported and used,
1386 regardless of this setting.
1388 Default value: \fB1,048,576\fR.
1394 \fBzfs_mdcomp_disable\fR (int)
1397 Disable meta data compression
1399 Use \fB1\fR for yes and \fB0\fR for no (default).
1405 \fBzfs_metaslab_fragmentation_threshold\fR (int)
1408 Allow metaslabs to keep their active state as long as their fragmentation
1409 percentage is less than or equal to this value. An active metaslab that
1410 exceeds this threshold will no longer keep its active status allowing
1411 better metaslabs to be selected.
1413 Default value: \fB70\fR.
1419 \fBzfs_mg_fragmentation_threshold\fR (int)
1422 Metaslab groups are considered eligible for allocations if their
1423 fragmentation metric (measured as a percentage) is less than or equal to
1424 this value. If a metaslab group exceeds this threshold then it will be
1425 skipped unless all metaslab groups within the metaslab class have also
1426 crossed this threshold.
1428 Default value: \fB85\fR.
1434 \fBzfs_mg_noalloc_threshold\fR (int)
1437 Defines a threshold at which metaslab groups should be eligible for
1438 allocations. The value is expressed as a percentage of free space
1439 beyond which a metaslab group is always eligible for allocations.
1440 If a metaslab group's free space is less than or equal to the
1441 threshold, the allocator will avoid allocating to that group
1442 unless all groups in the pool have reached the threshold. Once all
1443 groups have reached the threshold, all groups are allowed to accept
1444 allocations. The default value of 0 disables the feature and causes
1445 all metaslab groups to be eligible for allocations.
1447 This parameter allows one to deal with pools having heavily imbalanced
1448 vdevs such as would be the case when a new vdev has been added.
1449 Setting the threshold to a non-zero percentage will stop allocations
1450 from being made to vdevs that aren't filled to the specified percentage
1451 and allow lesser filled vdevs to acquire more allocations than they
1452 otherwise would under the old \fBzfs_mg_alloc_failures\fR facility.
1454 Default value: \fB0\fR.
1460 \fBzfs_multihost_history\fR (int)
1463 Historical statistics for the last N multihost updates will be available in
1464 \fB/proc/spl/kstat/zfs/<pool>/multihost\fR
1466 Default value: \fB0\fR.
1472 \fBzfs_multihost_interval\fR (ulong)
1475 Used to control the frequency of multihost writes which are performed when the
1476 \fBmultihost\fR pool property is on. This is one factor used to determine
1477 the length of the activity check during import.
1479 The multihost write period is \fBzfs_multihost_interval / leaf-vdevs\fR milliseconds.
1480 This means that on average a multihost write will be issued for each leaf vdev every
1481 \fBzfs_multihost_interval\fR milliseconds. In practice, the observed period can
1482 vary with the I/O load and this observed value is the delay which is stored in
1485 On import the activity check waits a minimum amount of time determined by
1486 \fBzfs_multihost_interval * zfs_multihost_import_intervals\fR. The activity
1487 check time may be further extended if the value of mmp delay found in the best
1488 uberblock indicates actual multihost updates happened at longer intervals than
1489 \fBzfs_multihost_interval\fR. A minimum value of \fB100ms\fR is enforced.
1491 Default value: \fB1000\fR.
1497 \fBzfs_multihost_import_intervals\fR (uint)
1500 Used to control the duration of the activity test on import. Smaller values of
1501 \fBzfs_multihost_import_intervals\fR will reduce the import time but increase
1502 the risk of failing to detect an active pool. The total activity check time is
1503 never allowed to drop below one second. A value of 0 is ignored and treated as
1506 Default value: \fB10\fR.
1512 \fBzfs_multihost_fail_intervals\fR (uint)
1515 Controls the behavior of the pool when multihost write failures are detected.
1517 When \fBzfs_multihost_fail_intervals = 0\fR then multihost write failures are ignored.
1518 The failures will still be reported to the ZED which depending on its
1519 configuration may take action such as suspending the pool or offlining a device.
1521 When \fBzfs_multihost_fail_intervals > 0\fR then sequential multihost write failures
1522 will cause the pool to be suspended. This occurs when
1523 \fBzfs_multihost_fail_intervals * zfs_multihost_interval\fR milliseconds have
1524 passed since the last successful multihost write. This guarantees the activity test
1525 will see multihost writes if the pool is imported.
1527 Default value: \fB5\fR.
1533 \fBzfs_no_scrub_io\fR (int)
1536 Set for no scrub I/O. This results in scrubs not actually scrubbing data and
1537 simply doing a metadata crawl of the pool instead.
1539 Use \fB1\fR for yes and \fB0\fR for no (default).
1545 \fBzfs_no_scrub_prefetch\fR (int)
1548 Set to disable block prefetching for scrubs.
1550 Use \fB1\fR for yes and \fB0\fR for no (default).
1556 \fBzfs_nocacheflush\fR (int)
1559 Disable cache flush operations on disks when writing. Beware, this may cause
1560 corruption if disks re-order writes.
1562 Use \fB1\fR for yes and \fB0\fR for no (default).
1568 \fBzfs_nopwrite_enabled\fR (int)
1573 Use \fB1\fR for yes (default) and \fB0\fR to disable.
1579 \fBzfs_dmu_offset_next_sync\fR (int)
1582 Enable forcing txg sync to find holes. When enabled forces ZFS to act
1583 like prior versions when SEEK_HOLE or SEEK_DATA flags are used, which
1584 when a dnode is dirty causes txg's to be synced so that this data can be
1587 Use \fB1\fR for yes and \fB0\fR to disable (default).
1593 \fBzfs_pd_bytes_max\fR (int)
1596 The number of bytes which should be prefetched during a pool traversal
1597 (eg: \fBzfs send\fR or other data crawling operations)
1599 Default value: \fB52,428,800\fR.
1605 \fBzfs_per_txg_dirty_frees_percent \fR (ulong)
1608 Tunable to control percentage of dirtied blocks from frees in one TXG.
1609 After this threshold is crossed, additional dirty blocks from frees
1610 wait until the next TXG.
1611 A value of zero will disable this throttle.
1613 Default value: \fB30\fR and \fB0\fR to disable.
1621 \fBzfs_prefetch_disable\fR (int)
1624 This tunable disables predictive prefetch. Note that it leaves "prescient"
1625 prefetch (e.g. prefetch for zfs send) intact. Unlike predictive prefetch,
1626 prescient prefetch never issues i/os that end up not being needed, so it
1627 can't hurt performance.
1629 Use \fB1\fR for yes and \fB0\fR for no (default).
1635 \fBzfs_read_chunk_size\fR (long)
1638 Bytes to read per chunk
1640 Default value: \fB1,048,576\fR.
1646 \fBzfs_read_history\fR (int)
1649 Historical statistics for the last N reads will be available in
1650 \fB/proc/spl/kstat/zfs/<pool>/reads\fR
1652 Default value: \fB0\fR (no data is kept).
1658 \fBzfs_read_history_hits\fR (int)
1661 Include cache hits in read history
1663 Use \fB1\fR for yes and \fB0\fR for no (default).
1669 \fBzfs_recover\fR (int)
1672 Set to attempt to recover from fatal errors. This should only be used as a
1673 last resort, as it typically results in leaked space, or worse.
1675 Use \fB1\fR for yes and \fB0\fR for no (default).
1681 \fBzfs_resilver_min_time_ms\fR (int)
1684 Resilvers are processed by the sync thread. While resilvering it will spend
1685 at least this much time working on a resilver between txg flushes.
1687 Default value: \fB3,000\fR.
1693 \fBzfs_scrub_min_time_ms\fR (int)
1696 Scrubs are processed by the sync thread. While scrubbing it will spend
1697 at least this much time working on a scrub between txg flushes.
1699 Default value: \fB1,000\fR.
1705 \fBzfs_scan_checkpoint_intval\fR (int)
1708 To preserve progress across reboots the sequential scan algorithm periodically
1709 needs to stop metadata scanning and issue all the verifications I/Os to disk.
1710 The frequency of this flushing is determined by the
1711 \fBfBzfs_scan_checkpoint_intval\fR tunable.
1713 Default value: \fB7200\fR seconds (every 2 hours).
1719 \fBzfs_scan_fill_weight\fR (int)
1722 This tunable affects how scrub and resilver I/O segments are ordered. A higher
1723 number indicates that we care more about how filled in a segment is, while a
1724 lower number indicates we care more about the size of the extent without
1725 considering the gaps within a segment. This value is only tunable upon module
1726 insertion. Changing the value afterwards will have no affect on scrub or
1727 resilver performance.
1729 Default value: \fB3\fR.
1735 \fBzfs_scan_issue_strategy\fR (int)
1738 Determines the order that data will be verified while scrubbing or resilvering.
1739 If set to \fB1\fR, data will be verified as sequentially as possible, given the
1740 amount of memory reserved for scrubbing (see \fBzfs_scan_mem_lim_fact\fR). This
1741 may improve scrub performance if the pool's data is very fragmented. If set to
1742 \fB2\fR, the largest mostly-contiguous chunk of found data will be verified
1743 first. By deferring scrubbing of small segments, we may later find adjacent data
1744 to coalesce and increase the segment size. If set to \fB0\fR, zfs will use
1745 strategy \fB1\fR during normal verification and strategy \fB2\fR while taking a
1748 Default value: \fB0\fR.
1754 \fBzfs_scan_legacy\fR (int)
1757 A value of 0 indicates that scrubs and resilvers will gather metadata in
1758 memory before issuing sequential I/O. A value of 1 indicates that the legacy
1759 algorithm will be used where I/O is initiated as soon as it is discovered.
1760 Changing this value to 0 will not affect scrubs or resilvers that are already
1763 Default value: \fB0\fR.
1769 \fBzfs_scan_max_ext_gap\fR (int)
1772 Indicates the largest gap in bytes between scrub / resilver I/Os that will still
1773 be considered sequential for sorting purposes. Changing this value will not
1774 affect scrubs or resilvers that are already in progress.
1776 Default value: \fB2097152 (2 MB)\fR.
1782 \fBzfs_scan_mem_lim_fact\fR (int)
1785 Maximum fraction of RAM used for I/O sorting by sequential scan algorithm.
1786 This tunable determines the hard limit for I/O sorting memory usage.
1787 When the hard limit is reached we stop scanning metadata and start issuing
1788 data verification I/O. This is done until we get below the soft limit.
1790 Default value: \fB20\fR which is 5% of RAM (1/20).
1796 \fBzfs_scan_mem_lim_soft_fact\fR (int)
1799 The fraction of the hard limit used to determined the soft limit for I/O sorting
1800 by the sequential scan algorithm. When we cross this limit from bellow no action
1801 is taken. When we cross this limit from above it is because we are issuing
1802 verification I/O. In this case (unless the metadata scan is done) we stop
1803 issuing verification I/O and start scanning metadata again until we get to the
1806 Default value: \fB20\fR which is 5% of the hard limit (1/20).
1812 \fBzfs_scan_vdev_limit\fR (int)
1815 Maximum amount of data that can be concurrently issued at once for scrubs and
1816 resilvers per leaf device, given in bytes.
1818 Default value: \fB41943040\fR.
1824 \fBzfs_send_corrupt_data\fR (int)
1827 Allow sending of corrupt data (ignore read/checksum errors when sending data)
1829 Use \fB1\fR for yes and \fB0\fR for no (default).
1835 \fBzfs_sync_pass_deferred_free\fR (int)
1838 Flushing of data to disk is done in passes. Defer frees starting in this pass
1840 Default value: \fB2\fR.
1846 \fBzfs_sync_pass_dont_compress\fR (int)
1849 Don't compress starting in this pass
1851 Default value: \fB5\fR.
1857 \fBzfs_sync_pass_rewrite\fR (int)
1860 Rewrite new block pointers starting in this pass
1862 Default value: \fB2\fR.
1868 \fBzfs_sync_taskq_batch_pct\fR (int)
1871 This controls the number of threads used by the dp_sync_taskq. The default
1872 value of 75% will create a maximum of one thread per cpu.
1874 Default value: \fB75\fR%.
1880 \fBzfs_txg_history\fR (int)
1883 Historical statistics for the last N txgs will be available in
1884 \fB/proc/spl/kstat/zfs/<pool>/txgs\fR
1886 Default value: \fB0\fR.
1892 \fBzfs_txg_timeout\fR (int)
1895 Flush dirty data to disk at least every N seconds (maximum txg duration)
1897 Default value: \fB5\fR.
1903 \fBzfs_vdev_aggregation_limit\fR (int)
1906 Max vdev I/O aggregation size
1908 Default value: \fB131,072\fR.
1914 \fBzfs_vdev_cache_bshift\fR (int)
1917 Shift size to inflate reads too
1919 Default value: \fB16\fR (effectively 65536).
1925 \fBzfs_vdev_cache_max\fR (int)
1928 Inflate reads smaller than this value to meet the \fBzfs_vdev_cache_bshift\fR
1931 Default value: \fB16384\fR.
1937 \fBzfs_vdev_cache_size\fR (int)
1940 Total size of the per-disk cache in bytes.
1942 Currently this feature is disabled as it has been found to not be helpful
1943 for performance and in some cases harmful.
1945 Default value: \fB0\fR.
1951 \fBzfs_vdev_mirror_rotating_inc\fR (int)
1954 A number by which the balancing algorithm increments the load calculation for
1955 the purpose of selecting the least busy mirror member when an I/O immediately
1956 follows its predecessor on rotational vdevs for the purpose of making decisions
1959 Default value: \fB0\fR.
1965 \fBzfs_vdev_mirror_rotating_seek_inc\fR (int)
1968 A number by which the balancing algorithm increments the load calculation for
1969 the purpose of selecting the least busy mirror member when an I/O lacks
1970 locality as defined by the zfs_vdev_mirror_rotating_seek_offset. I/Os within
1971 this that are not immediately following the previous I/O are incremented by
1974 Default value: \fB5\fR.
1980 \fBzfs_vdev_mirror_rotating_seek_offset\fR (int)
1983 The maximum distance for the last queued I/O in which the balancing algorithm
1984 considers an I/O to have locality.
1985 See the section "ZFS I/O SCHEDULER".
1987 Default value: \fB1048576\fR.
1993 \fBzfs_vdev_mirror_non_rotating_inc\fR (int)
1996 A number by which the balancing algorithm increments the load calculation for
1997 the purpose of selecting the least busy mirror member on non-rotational vdevs
1998 when I/Os do not immediately follow one another.
2000 Default value: \fB0\fR.
2006 \fBzfs_vdev_mirror_non_rotating_seek_inc\fR (int)
2009 A number by which the balancing algorithm increments the load calculation for
2010 the purpose of selecting the least busy mirror member when an I/O lacks
2011 locality as defined by the zfs_vdev_mirror_rotating_seek_offset. I/Os within
2012 this that are not immediately following the previous I/O are incremented by
2015 Default value: \fB1\fR.
2021 \fBzfs_vdev_read_gap_limit\fR (int)
2024 Aggregate read I/O operations if the gap on-disk between them is within this
2027 Default value: \fB32,768\fR.
2033 \fBzfs_vdev_scheduler\fR (charp)
2036 Set the Linux I/O scheduler on whole disk vdevs to this scheduler. Valid options
2037 are noop, cfq, bfq & deadline
2039 Default value: \fBnoop\fR.
2045 \fBzfs_vdev_write_gap_limit\fR (int)
2048 Aggregate write I/O over gap
2050 Default value: \fB4,096\fR.
2056 \fBzfs_vdev_raidz_impl\fR (string)
2059 Parameter for selecting raidz parity implementation to use.
2061 Options marked (always) below may be selected on module load as they are
2062 supported on all systems.
2063 The remaining options may only be set after the module is loaded, as they
2064 are available only if the implementations are compiled in and supported
2065 on the running system.
2067 Once the module is loaded, the content of
2068 /sys/module/zfs/parameters/zfs_vdev_raidz_impl will show available options
2069 with the currently selected one enclosed in [].
2070 Possible options are:
2071 fastest - (always) implementation selected using built-in benchmark
2072 original - (always) original raidz implementation
2073 scalar - (always) scalar raidz implementation
2074 sse2 - implementation using SSE2 instruction set (64bit x86 only)
2075 ssse3 - implementation using SSSE3 instruction set (64bit x86 only)
2076 avx2 - implementation using AVX2 instruction set (64bit x86 only)
2077 avx512f - implementation using AVX512F instruction set (64bit x86 only)
2078 avx512bw - implementation using AVX512F & AVX512BW instruction sets (64bit x86 only)
2079 aarch64_neon - implementation using NEON (Aarch64/64 bit ARMv8 only)
2080 aarch64_neonx2 - implementation using NEON with more unrolling (Aarch64/64 bit ARMv8 only)
2082 Default value: \fBfastest\fR.
2088 \fBzfs_zevent_cols\fR (int)
2091 When zevents are logged to the console use this as the word wrap width.
2093 Default value: \fB80\fR.
2099 \fBzfs_zevent_console\fR (int)
2102 Log events to the console
2104 Use \fB1\fR for yes and \fB0\fR for no (default).
2110 \fBzfs_zevent_len_max\fR (int)
2113 Max event queue length. A value of 0 will result in a calculated value which
2114 increases with the number of CPUs in the system (minimum 64 events). Events
2115 in the queue can be viewed with the \fBzpool events\fR command.
2117 Default value: \fB0\fR.
2123 \fBzfs_zil_clean_taskq_maxalloc\fR (int)
2126 The maximum number of taskq entries that are allowed to be cached. When this
2127 limit is exceeded transaction records (itxs) will be cleaned synchronously.
2129 Default value: \fB1048576\fR.
2135 \fBzfs_zil_clean_taskq_minalloc\fR (int)
2138 The number of taskq entries that are pre-populated when the taskq is first
2139 created and are immediately available for use.
2141 Default value: \fB1024\fR.
2147 \fBzfs_zil_clean_taskq_nthr_pct\fR (int)
2150 This controls the number of threads used by the dp_zil_clean_taskq. The default
2151 value of 100% will create a maximum of one thread per cpu.
2153 Default value: \fB100\fR%.
2159 \fBzil_replay_disable\fR (int)
2162 Disable intent logging replay. Can be disabled for recovery from corrupted
2165 Use \fB1\fR for yes and \fB0\fR for no (default).
2171 \fBzil_slog_bulk\fR (ulong)
2174 Limit SLOG write size per commit executed with synchronous priority.
2175 Any writes above that will be executed with lower (asynchronous) priority
2176 to limit potential SLOG device abuse by single active ZIL writer.
2178 Default value: \fB786,432\fR.
2184 \fBzio_delay_max\fR (int)
2187 A zevent will be logged if a ZIO operation takes more than N milliseconds to
2188 complete. Note that this is only a logging facility, not a timeout on
2191 Default value: \fB30,000\fR.
2197 \fBzio_dva_throttle_enabled\fR (int)
2200 Throttle block allocations in the ZIO pipeline. This allows for
2201 dynamic allocation distribution when devices are imbalanced.
2202 When enabled, the maximum number of pending allocations per top-level vdev
2203 is limited by \fBzfs_vdev_queue_depth_pct\fR.
2205 Default value: \fB1\fR.
2211 \fBzio_requeue_io_start_cut_in_line\fR (int)
2214 Prioritize requeued I/O
2216 Default value: \fB0\fR.
2222 \fBzio_taskq_batch_pct\fR (uint)
2225 Percentage of online CPUs (or CPU cores, etc) which will run a worker thread
2226 for IO. These workers are responsible for IO work such as compression and
2227 checksum calculations. Fractional number of CPUs will be rounded down.
2229 The default value of 75 was chosen to avoid using all CPUs which can result in
2230 latency issues and inconsistent application performance, especially when high
2231 compression is enabled.
2233 Default value: \fB75\fR.
2239 \fBzvol_inhibit_dev\fR (uint)
2242 Do not create zvol device nodes. This may slightly improve startup time on
2243 systems with a very large number of zvols.
2245 Use \fB1\fR for yes and \fB0\fR for no (default).
2251 \fBzvol_major\fR (uint)
2254 Major number for zvol block devices
2256 Default value: \fB230\fR.
2262 \fBzvol_max_discard_blocks\fR (ulong)
2265 Discard (aka TRIM) operations done on zvols will be done in batches of this
2266 many blocks, where block size is determined by the \fBvolblocksize\fR property
2269 Default value: \fB16,384\fR.
2275 \fBzvol_prefetch_bytes\fR (uint)
2278 When adding a zvol to the system prefetch \fBzvol_prefetch_bytes\fR
2279 from the start and end of the volume. Prefetching these regions
2280 of the volume is desirable because they are likely to be accessed
2281 immediately by \fBblkid(8)\fR or by the kernel scanning for a partition
2284 Default value: \fB131,072\fR.
2290 \fBzvol_request_sync\fR (uint)
2293 When processing I/O requests for a zvol submit them synchronously. This
2294 effectively limits the queue depth to 1 for each I/O submitter. When set
2295 to 0 requests are handled asynchronously by a thread pool. The number of
2296 requests which can be handled concurrently is controller by \fBzvol_threads\fR.
2298 Default value: \fB0\fR.
2304 \fBzvol_threads\fR (uint)
2307 Max number of threads which can handle zvol I/O requests concurrently.
2309 Default value: \fB32\fR.
2315 \fBzvol_volmode\fR (uint)
2318 Defines zvol block devices behaviour when \fBvolmode\fR is set to \fBdefault\fR.
2319 Valid values are \fB1\fR (full), \fB2\fR (dev) and \fB3\fR (none).
2321 Default value: \fB1\fR.
2327 \fBzfs_qat_disable\fR (int)
2330 This tunable disables qat hardware acceleration for gzip compression.
2331 It is available only if qat acceleration is compiled in and qat driver
2334 Use \fB1\fR for yes and \fB0\fR for no (default).
2337 .SH ZFS I/O SCHEDULER
2338 ZFS issues I/O operations to leaf vdevs to satisfy and complete I/Os.
2339 The I/O scheduler determines when and in what order those operations are
2340 issued. The I/O scheduler divides operations into five I/O classes
2341 prioritized in the following order: sync read, sync write, async read,
2342 async write, and scrub/resilver. Each queue defines the minimum and
2343 maximum number of concurrent operations that may be issued to the
2344 device. In addition, the device has an aggregate maximum,
2345 \fBzfs_vdev_max_active\fR. Note that the sum of the per-queue minimums
2346 must not exceed the aggregate maximum. If the sum of the per-queue
2347 maximums exceeds the aggregate maximum, then the number of active I/Os
2348 may reach \fBzfs_vdev_max_active\fR, in which case no further I/Os will
2349 be issued regardless of whether all per-queue minimums have been met.
2351 For many physical devices, throughput increases with the number of
2352 concurrent operations, but latency typically suffers. Further, physical
2353 devices typically have a limit at which more concurrent operations have no
2354 effect on throughput or can actually cause it to decrease.
2356 The scheduler selects the next operation to issue by first looking for an
2357 I/O class whose minimum has not been satisfied. Once all are satisfied and
2358 the aggregate maximum has not been hit, the scheduler looks for classes
2359 whose maximum has not been satisfied. Iteration through the I/O classes is
2360 done in the order specified above. No further operations are issued if the
2361 aggregate maximum number of concurrent operations has been hit or if there
2362 are no operations queued for an I/O class that has not hit its maximum.
2363 Every time an I/O is queued or an operation completes, the I/O scheduler
2364 looks for new operations to issue.
2366 In general, smaller max_active's will lead to lower latency of synchronous
2367 operations. Larger max_active's may lead to higher overall throughput,
2368 depending on underlying storage.
2370 The ratio of the queues' max_actives determines the balance of performance
2371 between reads, writes, and scrubs. E.g., increasing
2372 \fBzfs_vdev_scrub_max_active\fR will cause the scrub or resilver to complete
2373 more quickly, but reads and writes to have higher latency and lower throughput.
2375 All I/O classes have a fixed maximum number of outstanding operations
2376 except for the async write class. Asynchronous writes represent the data
2377 that is committed to stable storage during the syncing stage for
2378 transaction groups. Transaction groups enter the syncing state
2379 periodically so the number of queued async writes will quickly burst up
2380 and then bleed down to zero. Rather than servicing them as quickly as
2381 possible, the I/O scheduler changes the maximum number of active async
2382 write I/Os according to the amount of dirty data in the pool. Since
2383 both throughput and latency typically increase with the number of
2384 concurrent operations issued to physical devices, reducing the
2385 burstiness in the number of concurrent operations also stabilizes the
2386 response time of operations from other -- and in particular synchronous
2387 -- queues. In broad strokes, the I/O scheduler will issue more
2388 concurrent operations from the async write queue as there's more dirty
2393 The number of concurrent operations issued for the async write I/O class
2394 follows a piece-wise linear function defined by a few adjustable points.
2397 | o---------| <-- zfs_vdev_async_write_max_active
2404 |-------o | | <-- zfs_vdev_async_write_min_active
2405 0|_______^______|_________|
2406 0% | | 100% of zfs_dirty_data_max
2408 | `-- zfs_vdev_async_write_active_max_dirty_percent
2409 `--------- zfs_vdev_async_write_active_min_dirty_percent
2412 Until the amount of dirty data exceeds a minimum percentage of the dirty
2413 data allowed in the pool, the I/O scheduler will limit the number of
2414 concurrent operations to the minimum. As that threshold is crossed, the
2415 number of concurrent operations issued increases linearly to the maximum at
2416 the specified maximum percentage of the dirty data allowed in the pool.
2418 Ideally, the amount of dirty data on a busy pool will stay in the sloped
2419 part of the function between \fBzfs_vdev_async_write_active_min_dirty_percent\fR
2420 and \fBzfs_vdev_async_write_active_max_dirty_percent\fR. If it exceeds the
2421 maximum percentage, this indicates that the rate of incoming data is
2422 greater than the rate that the backend storage can handle. In this case, we
2423 must further throttle incoming writes, as described in the next section.
2425 .SH ZFS TRANSACTION DELAY
2426 We delay transactions when we've determined that the backend storage
2427 isn't able to accommodate the rate of incoming writes.
2429 If there is already a transaction waiting, we delay relative to when
2430 that transaction will finish waiting. This way the calculated delay time
2431 is independent of the number of threads concurrently executing
2434 If we are the only waiter, wait relative to when the transaction
2435 started, rather than the current time. This credits the transaction for
2436 "time already served", e.g. reading indirect blocks.
2438 The minimum time for a transaction to take is calculated as:
2440 min_time = zfs_delay_scale * (dirty - min) / (max - dirty)
2441 min_time is then capped at 100 milliseconds.
2444 The delay has two degrees of freedom that can be adjusted via tunables. The
2445 percentage of dirty data at which we start to delay is defined by
2446 \fBzfs_delay_min_dirty_percent\fR. This should typically be at or above
2447 \fBzfs_vdev_async_write_active_max_dirty_percent\fR so that we only start to
2448 delay after writing at full speed has failed to keep up with the incoming write
2449 rate. The scale of the curve is defined by \fBzfs_delay_scale\fR. Roughly speaking,
2450 this variable determines the amount of delay at the midpoint of the curve.
2454 10ms +-------------------------------------------------------------*+
2470 2ms + (midpoint) * +
2473 | zfs_delay_scale ----------> ******** |
2474 0 +-------------------------------------*********----------------+
2475 0% <- zfs_dirty_data_max -> 100%
2478 Note that since the delay is added to the outstanding time remaining on the
2479 most recent transaction, the delay is effectively the inverse of IOPS.
2480 Here the midpoint of 500us translates to 2000 IOPS. The shape of the curve
2481 was chosen such that small changes in the amount of accumulated dirty data
2482 in the first 3/4 of the curve yield relatively small differences in the
2485 The effects can be easier to understand when the amount of delay is
2486 represented on a log scale:
2490 100ms +-------------------------------------------------------------++
2499 + zfs_delay_scale ----------> ***** +
2510 +--------------------------------------------------------------+
2511 0% <- zfs_dirty_data_max -> 100%
2514 Note here that only as the amount of dirty data approaches its limit does
2515 the delay start to increase rapidly. The goal of a properly tuned system
2516 should be to keep the amount of dirty data out of that range by first
2517 ensuring that the appropriate limits are set for the I/O scheduler to reach
2518 optimal throughput on the backend storage, and then by changing the value
2519 of \fBzfs_delay_scale\fR to increase the steepness of the curve.
projects / zfs / blob