The policy can return a simple HIT or MISS or issue a migration.
Currently there's no way for the policy to issue background work,
-e.g. to start writing back dirty blocks that are going to be evicte
+e.g. to start writing back dirty blocks that are going to be evicted
soon.
Because we map bios, rather than requests it's easy for the policy
Overview of supplied cache replacement policies
===============================================
-multiqueue
-----------
+multiqueue (mq)
+---------------
-This policy is the default.
-
-The multiqueue policy has three sets of 16 queues: one set for entries
-waiting for the cache and another two for those in the cache (a set for
-clean entries and a set for dirty entries).
+This policy is now an alias for smq (see below).
-Cache entries in the queues are aged based on logical time. Entry into
-the cache is based on variable thresholds and queue selection is based
-on hit count on entry. The policy aims to take different cache miss
-costs into account and to adjust to varying load patterns automatically.
+The following tunables are accepted, but have no effect:
-Message and constructor argument pairs are:
'sequential_threshold <#nr_sequential_ios>'
'random_threshold <#nr_random_ios>'
'read_promote_adjustment <value>'
'write_promote_adjustment <value>'
'discard_promote_adjustment <value>'
-The sequential threshold indicates the number of contiguous I/Os
-required before a stream is treated as sequential. Once a stream is
-considered sequential it will bypass the cache. The random threshold
-is the number of intervening non-contiguous I/Os that must be seen
-before the stream is treated as random again.
-
-The sequential and random thresholds default to 512 and 4 respectively.
-
-Large, sequential I/Os are probably better left on the origin device
-since spindles tend to have good sequential I/O bandwidth. The
-io_tracker counts contiguous I/Os to try to spot when the I/O is in one
-of these sequential modes. But there are use-cases for wanting to
-promote sequential blocks to the cache (e.g. fast application startup).
-If sequential threshold is set to 0 the sequential I/O detection is
-disabled and sequential I/O will no longer implicitly bypass the cache.
-Setting the random threshold to 0 does _not_ disable the random I/O
-stream detection.
-
-Internally the mq policy determines a promotion threshold. If the hit
-count of a block not in the cache goes above this threshold it gets
-promoted to the cache. The read, write and discard promote adjustment
-tunables allow you to tweak the promotion threshold by adding a small
-value based on the io type. They default to 4, 8 and 1 respectively.
-If you're trying to quickly warm a new cache device you may wish to
-reduce these to encourage promotion. Remember to switch them back to
-their defaults after the cache fills though.
+Stochastic multiqueue (smq)
+---------------------------
+
+This policy is the default.
+
+The stochastic multi-queue (smq) policy addresses some of the problems
+with the multiqueue (mq) policy.
+
+The smq policy (vs mq) offers the promise of less memory utilization,
+improved performance and increased adaptability in the face of changing
+workloads. smq also does not have any cumbersome tuning knobs.
+
+Users may switch from "mq" to "smq" simply by appropriately reloading a
+DM table that is using the cache target. Doing so will cause all of the
+mq policy's hints to be dropped. Also, performance of the cache may
+degrade slightly until smq recalculates the origin device's hotspots
+that should be cached.
+
+Memory usage:
+The mq policy used a lot of memory; 88 bytes per cache block on a 64
+bit machine.
+
+smq uses 28bit indexes to implement it's data structures rather than
+pointers. It avoids storing an explicit hit count for each block. It
+has a 'hotspot' queue, rather than a pre-cache, which uses a quarter of
+the entries (each hotspot block covers a larger area than a single
+cache block).
+
+All this means smq uses ~25bytes per cache block. Still a lot of
+memory, but a substantial improvement nontheless.
+
+Level balancing:
+mq placed entries in different levels of the multiqueue structures
+based on their hit count (~ln(hit count)). This meant the bottom
+levels generally had the most entries, and the top ones had very
+few. Having unbalanced levels like this reduced the efficacy of the
+multiqueue.
+
+smq does not maintain a hit count, instead it swaps hit entries with
+the least recently used entry from the level above. The overall
+ordering being a side effect of this stochastic process. With this
+scheme we can decide how many entries occupy each multiqueue level,
+resulting in better promotion/demotion decisions.
+
+Adaptability:
+The mq policy maintained a hit count for each cache block. For a
+different block to get promoted to the cache it's hit count has to
+exceed the lowest currently in the cache. This meant it could take a
+long time for the cache to adapt between varying IO patterns.
+
+smq doesn't maintain hit counts, so a lot of this problem just goes
+away. In addition it tracks performance of the hotspot queue, which
+is used to decide which blocks to promote. If the hotspot queue is
+performing badly then it starts moving entries more quickly between
+levels. This lets it adapt to new IO patterns very quickly.
+
+Performance:
+Testing smq shows substantially better performance than mq.
cleaner
-------
<#read hits> <#read misses> <#write hits> <#write misses>
<#demotions> <#promotions> <#dirty> <#features> <features>*
<#core args> <core args>* <policy name> <#policy args> <policy args>*
+<cache metadata mode>
metadata block size : Fixed block size for each metadata block in
sectors
e.g. migration_threshold
policy name : Name of the policy
#policy args : Number of policy arguments to follow (must be even)
-policy args : Key/value pairs
- e.g. sequential_threshold
+policy args : Key/value pairs e.g. sequential_threshold
+cache metadata mode : ro if read-only, rw if read-write
+ In serious cases where even a read-only mode is deemed unsafe
+ no further I/O will be permitted and the status will just
+ contain the string 'Fail'. The userspace recovery tools
+ should then be used.
+needs_check : 'needs_check' if set, '-' if not set
+ A metadata operation has failed, resulting in the needs_check
+ flag being set in the metadata's superblock. The metadata
+ device must be deactivated and checked/repaired before the
+ cache can be made fully operational again. '-' indicates
+ needs_check is not set.
Messages
--------
<device> <offset> <delay> [<write_device> <write_offset> <write_delay>]
With separate write parameters, the first set is only used for reads.
+Offsets are specified in sectors.
Delays are specified in milliseconds.
Example scripts
"repair" - Initiate a repair of the array.
"reshape"- Currently unsupported (-EINVAL).
+
+Discard Support
+---------------
+The implementation of discard support among hardware vendors varies.
+When a block is discarded, some storage devices will return zeroes when
+the block is read. These devices set the 'discard_zeroes_data'
+attribute. Other devices will return random data. Confusingly, some
+devices that advertise 'discard_zeroes_data' will not reliably return
+zeroes when discarded blocks are read! Since RAID 4/5/6 uses blocks
+from a number of devices to calculate parity blocks and (for performance
+reasons) relies on 'discard_zeroes_data' being reliable, it is important
+that the devices be consistent. Blocks may be discarded in the middle
+of a RAID 4/5/6 stripe and if subsequent read results are not
+consistent, the parity blocks may be calculated differently at any time;
+making the parity blocks useless for redundancy. It is important to
+understand how your hardware behaves with discards if you are going to
+enable discards with RAID 4/5/6.
+
+Since the behavior of storage devices is unreliable in this respect,
+even when reporting 'discard_zeroes_data', by default RAID 4/5/6
+discard support is disabled -- this ensures data integrity at the
+expense of losing some performance.
+
+Storage devices that properly support 'discard_zeroes_data' are
+increasingly whitelisted in the kernel and can thus be trusted.
+
+For trusted devices, the following dm-raid module parameter can be set
+to safely enable discard support for RAID 4/5/6:
+ 'devices_handle_discards_safely'
+
+
Version History
---------------
1.0.0 Initial version. Support for RAID 4/5/6
New status (STATUSTYPE_INFO) fields: sync_action and mismatch_cnt.
1.5.1 Add ability to restore transiently failed devices on resume.
1.5.2 'mismatch_cnt' is zero unless [last_]sync_action is "check".
+1.6.0 Add discard support (and devices_handle_discard_safely module param).
+1.7.0 Add support for MD RAID0 mappings.
the amount of free space and expand the <COW device> before it fills up.
<persistent?> is P (Persistent) or N (Not persistent - will not survive
-after reboot).
-The difference is that for transient snapshots less metadata must be
-saved on disk - they can be kept in memory by the kernel.
+after reboot). O (Overflow) can be added as a persistent store option
+to allow userspace to advertise its support for seeing "Overflow" in the
+snapshot status. So supported store types are "P", "PO" and "N".
+
+The difference between persistent and transient is with transient
+snapshots less metadata must be saved on disk - they can be kept in
+memory by the kernel.
* snapshot-merge <origin> <COW device> <persistent> <chunksize>
The I/O statistics counters for each step-sized area of a region are
in the same format as /sys/block/*/stat or /proc/diskstats (see:
Documentation/iostats.txt). But two extra counters (12 and 13) are
-provided: total time spent reading and writing in milliseconds. All
-these counters may be accessed by sending the @stats_print message to
-the appropriate DM device via dmsetup.
+provided: total time spent reading and writing. When the histogram
+argument is used, the 14th parameter is reported that represents the
+histogram of latencies. All these counters may be accessed by sending
+the @stats_print message to the appropriate DM device via dmsetup.
+
+The reported times are in milliseconds and the granularity depends on
+the kernel ticks. When the option precise_timestamps is used, the
+reported times are in nanoseconds.
Each region has a corresponding unique identifier, which we call a
region_id, that is assigned when the region is created. The region_id
Messages
========
- @stats_create <range> <step> [<program_id> [<aux_data>]]
+ @stats_create <range> <step>
+ [<number_of_optional_arguments> <optional_arguments>...]
+ [<program_id> [<aux_data>]]
Create a new region and return the region_id.
"/<number_of_areas>" - the range is subdivided into the specified
number of areas.
+ <number_of_optional_arguments>
+ The number of optional arguments
+
+ <optional_arguments>
+ The following optional arguments are supported
+ precise_timestamps - use precise timer with nanosecond resolution
+ instead of the "jiffies" variable. When this argument is
+ used, the resulting times are in nanoseconds instead of
+ milliseconds. Precise timestamps are a little bit slower
+ to obtain than jiffies-based timestamps.
+ histogram:n1,n2,n3,n4,... - collect histogram of latencies. The
+ numbers n1, n2, etc are times that represent the boundaries
+ of the histogram. If precise_timestamps is not used, the
+ times are in milliseconds, otherwise they are in
+ nanoseconds. For each range, the kernel will report the
+ number of requests that completed within this range. For
+ example, if we use "histogram:10,20,30", the kernel will
+ report four numbers a:b:c:d. a is the number of requests
+ that took 0-10 ms to complete, b is the number of requests
+ that took 10-20 ms to complete, c is the number of requests
+ that took 20-30 ms to complete and d is the number of
+ requests that took more than 30 ms to complete.
+
<program_id>
An optional parameter. A name that uniquely identifies
the userspace owner of the range. This groups ranges together
created and ignore those created by others.
The kernel returns this string back in the output of
@stats_list message, but it doesn't use it for anything else.
+ If we omit the number of optional arguments, program id must not
+ be a number, otherwise it would be interpreted as the number of
+ optional arguments.
<aux_data>
An optional parameter. A word that provides auxiliary data
Output format:
<region_id>: <start_sector>+<length> <step> <program_id> <aux_data>
+ precise_timestamps histogram:n1,n2,n3,...
+
+ The strings "precise_timestamps" and "histogram" are printed only
+ if they were specified when creating the region.
@stats_print <region_id> [<starting_line> <number_of_lines>]
dmsetup message vol 0 @stats_create - /100
-Set the auxillary data string to "foo bar baz" (the escape for each
+Set the auxiliary data string to "foo bar baz" (the escape for each
space must also be escaped, otherwise the shell will consume them):
dmsetup message vol 0 @stats_set_aux 0 foo\\ bar\\ baz
underlying device. When this is enabled when loading the table,
it can get disabled if the underlying device doesn't support it.
- ro|rw
+ ro|rw|out_of_data_space
If the pool encounters certain types of device failures it will
drop into a read-only metadata mode in which no changes to
the pool metadata (like allocating new blocks) are permitted.
module parameter can be used to change this timeout -- it
defaults to 60 seconds but may be disabled using a value of 0.
+ needs_check
+ A metadata operation has failed, resulting in the needs_check
+ flag being set in the metadata's superblock. The metadata
+ device must be deactivated and checked/repaired before the
+ thin-pool can be made fully operational again. '-' indicates
+ needs_check is not set.
+
iii) Messages
create_thin <dev id>
0 is the original format used in the Chromium OS.
The salt is appended when hashing, digests are stored continuously and
- the rest of the block is padded with zeros.
+ the rest of the block is padded with zeroes.
1 is the current format that should be used for new devices.
The salt is prepended when hashing and each digest is
- padded with zeros to the power of two.
+ padded with zeroes to the power of two.
<dev>
This is the device containing data, the integrity of which needs to be
not compatible with ignore_corruption and requires user space support to
avoid restart loops.
+ignore_zero_blocks
+ Do not verify blocks that are expected to contain zeroes and always return
+ zeroes instead. This may be useful if the partition contains unused blocks
+ that are not guaranteed to contain zeroes.
+
+use_fec_from_device <fec_dev>
+ Use forward error correction (FEC) to recover from corruption if hash
+ verification fails. Use encoding data from the specified device. This
+ may be the same device where data and hash blocks reside, in which case
+ fec_start must be outside data and hash areas.
+
+ If the encoding data covers additional metadata, it must be accessible
+ on the hash device after the hash blocks.
+
+ Note: block sizes for data and hash devices must match. Also, if the
+ verity <dev> is encrypted the <fec_dev> should be too.
+
+fec_roots <num>
+ Number of generator roots. This equals to the number of parity bytes in
+ the encoding data. For example, in RS(M, N) encoding, the number of roots
+ is M-N.
+
+fec_blocks <num>
+ The number of encoding data blocks on the FEC device. The block size for
+ the FEC device is <data_block_size>.
+
+fec_start <offset>
+ This is the offset, in <data_block_size> blocks, from the start of the
+ FEC device to the beginning of the encoding data.
+
+
Theory of operation
===================
into the page cache. Block hashes are stored linearly, aligned to the nearest
block size.
+If forward error correction (FEC) support is enabled any recovery of
+corrupted data will be verified using the cryptographic hash of the
+corresponding data. This is why combining error correction with
+integrity checking is essential.
+
Hash Tree
---------