                        ===============================
                        SnapRAID Backup for Disk Arrays
                        ===============================


1 SYNOPSIS
==========

snapraid [-c, --conf CONFIG]
	[-f, --filter PATTERN] [-d, --filter-disk NAME]
	[-m, --filter-missing] [-e, --filter-error]
	[-a, --audit-only] [-h, --pre-hash] [-i, --import DIR]
	[-p, --plan PERC|bad|new|full]
	[-o, --older-than DAYS] [-l, --log FILE]
	[-s, --spin-down-on-error] [-w, --bw-limit RATE]
	[-t, --tail]
	[-Z, --force-zero] [-E, --force-empty]
	[-U, --force-uuid] [-D, --force-device]
	[-N, --force-nocopy] [-F, --force-full]
	[-R, --force-realloc] [-W, --force-realloc-tail]
	[-S, --start BLKSTART] [-B, --count BLKCOUNT]
	[-L, --error-limit NUMBER]
	[-A, --stats]
	[-v, --verbose] [-q, --quiet]
	status|smart|probe|up|down|diff|sync|scrub|fix|check
	|list|dup|pool|devices|touch|rehash|locate

snapraid [-V, --version] [-H, --help] [-C, --gen-conf CONTENT]


2 DESCRIPTION
=============

SnapRAID is a backup program designed for disk arrays, storing
parity information for data recovery in the event of up to six
disk failures.

Primarily intended for home media centers with large,
infrequently changing files, SnapRAID offers several features:

* You can utilize disks already filled with files without the
  need to reformat them, accessing them as usual.
* All your data is hashed to ensure data integrity and prevent
  silent corruption.
* When the number of failed disks exceeds the parity count,
  data loss is confined to the affected disks; data on
  other disks remains accessible.
* If you accidentally delete files on a disk, recovery is
  possible.
* Disks can have different sizes.
* You can add disks at any time.
* SnapRAID doesn't lock in your data; you can stop using it
  anytime without reformatting or moving data.
* To access a file, only a single disk needs to spin, saving
  power and reducing noise.

For more information, please visit the official SnapRAID site:

    https://www.snapraid.it/


3 LIMITATIONS
=============

SnapRAID is a hybrid between a RAID and a backup program, aiming to combine
the best benefits of both. However, it has some limitations that you should
consider before using it.

The main limitation is that if a disk fails, you can only recover data up to
the state of the last `sync` operation. Any data added or modifications made
since the last sync that are located on the failed disk will be lost.

Consequently, SnapRAID is primarily suited for data that rarely changes.

For data that already existed at the last sync, recovery reliability depends
on whether the `snapshot` option is used:

* With snapshot support (available on Btrfs, ZFS, Bcachefs, and NTFS),
  SnapRAID can automatically maintain an independent frozen reference
  of each supported data filesystem.
  This ensures that even if you modify or delete files on your live
  filesystem during or after a sync, the recovery process remains
  consistent and reliable.

* Without snapshot support, deleting or changing files after a `sync`
  can prevent the full recovery of other failed disks. This occurs
  because the parity no longer matches the modified files, even if
  those files are not on the failed disk. On the other hand, newly added
  files don't prevent recovery of already	existing files.

Other SnapRAID limitations are:

* With SnapRAID, you still have separate file systems for each disk.
  With RAID, you get a single large file system.
* SnapRAID doesn't stripe data.
  With RAID, you get a speed boost with striping.
* SnapRAID doesn't support real-time recovery.
  With RAID, you do not have to stop working when a disk fails.
* SnapRAID can recover data only from a limited number of disk failures.
  With a backup, you can recover from a complete
  failure of the entire disk array.
* Only file names, timestamps, symlinks, and hardlinks are saved.
  Permissions, ownership, and extended attributes are not saved.


4 GETTING STARTED
=================

To use SnapRAID, you need to first select one disk in your disk array
to dedicate to `parity` information. With one disk for parity, you
will be able to recover from a single disk failure, similar to RAID5.

If you want to recover from more disk failures, similar to RAID6,
you must reserve additional disks for parity. Each additional parity
disk allows recovery from one more disk failure.

As parity disks, you must pick the largest disks in the array,
as the parity information may grow to the size of the largest data
disk in the array.

These disks will be dedicated to storing the `parity` files.
You should not store your data on them.

Then, you must define the `data` disks that you want to protect
with SnapRAID. The protection is more effective if these disks
contain data that rarely changes. For this reason, it's better to
NOT include the Windows C:\ disk or the Unix /home, /var, and /tmp
directories.

The list of files is saved in the `content` files, usually
stored on the data, parity, or boot disks.
This file contains the details of your backup, including all the
checksums to verify its integrity.
The `content` file is stored in multiple copies, and each copy must
be on a different disk to ensure that, even in case of multiple
disk failures, at least one copy is available.

For example, suppose you are interested in only one parity level
of protection, and your disks are located at:

    /mnt/diskp <- selected disk for parity
    /mnt/disk1 <- first disk to protect
    /mnt/disk2 <- second disk to protect
    /mnt/disk3 <- third disk to protect

You must create the configuration file /etc/snapraid.conf with
the following options:

    parity /mnt/diskp/snapraid.parity
    content /var/snapraid/snapraid.content
    content /mnt/disk1/snapraid.content
    content /mnt/disk2/snapraid.content
    data d1 /mnt/disk1/
    data d2 /mnt/disk2/
    data d3 /mnt/disk3/

If you are on Windows, you should use the Windows path format, with drive
letters and backslashes instead of slashes.

    parity E:\snapraid.parity
    content C:\snapraid\snapraid.content
    content F:\array\snapraid.content
    content G:\array\snapraid.content
    data d1 F:\array\
    data d2 G:\array\
    data d3 H:\array\

If you have many disks and run out of drive letters, you can mount
disks directly in subfolders. See:

    https://www.google.com/search?q=Windows+mount+point

At this point, you are ready to run the `sync` command to build the
parity information.

    snapraid sync

This process may take several hours the first time, depending on the size
of the data already present on the disks. If the disks are empty,
the process is immediate.

You can stop it at any time by pressing Ctrl+C, and at the next run, it
will resume where it was interrupted.

When this command completes, your data is SAFE.

Now you can start using your array as you like and periodically
update the parity information by running the `sync` command.

4.1 Scrubbing
-------------

To periodically check the data and parity for errors, you can
run the `scrub` command.

    snapraid scrub

This command compares the data in your array with the hash computed
during the `sync` command to verify integrity.

Each run of the command checks approximately 8% of the array, excluding data
already scrubbed in the previous 10 days.
You can use the -p, --plan option to specify a different amount
and the -o, --older-than option to specify a different age in days.
For example, to check 5% of the array for blocks older than 20 days, use:

    snapraid -p 5 -o 20 scrub

If silent or input/output errors are found during the process,
the corresponding blocks are marked as bad in the `content` file
and listed in the `status` command.

    snapraid status

To fix them, you can use the `fix` command, filtering for bad blocks with
the -e, --filter-error option:

    snapraid -e fix

At the next `scrub`, the errors will disappear from the `status` report
if they are truly fixed. To make it faster, you can use -p bad to scrub
only blocks marked as bad.

    snapraid -p bad scrub

Running `scrub` on an unsynced array may report errors caused by
removed or modified files. These errors are reported in the `scrub`
output, but the related blocks are not marked as bad.

4.2 Pooling
-----------

Note: The pooling feature described below has been superseded by the
mergerfs tool, which is now the recommended option for Linux users in
the SnapRAID community. Mergerfs provides a more flexible and efficient
way to pool multiple drives into a single unified mount point,
allowing seamless access to files across your array without relying
on symbolic links. It integrates well with SnapRAID for parity
protection and is commonly used in setups like OpenMediaVault (OMV)
or custom NAS configurations.

To have all the files in your array shown in the same directory tree,
you can enable the `pooling` feature. It creates a read-only virtual
view of all the files in your array using symbolic links.

You can configure the `pooling` directory in the configuration file with:

    pool /pool

or, if you are on Windows, with:

    pool C:\pool

and then run the `pool` command to create or update the virtual view.

    snapraid pool

If you are using a Unix platform and want to share this directory
over the network to either Windows or Unix machines, you should add
the following options to your /etc/samba/smb.conf:

    # In the global section of smb.conf
    unix extensions = no

    # In the share section of smb.conf
    [pool]
    comment = Pool
    path = /pool
    read only = yes
    guest ok = yes
    wide links = yes
    follow symlinks = yes

In Windows, sharing symbolic links over a network requires clients to
resolve them remotely. To enable this, besides sharing the pool directory,
you must also share all the disks independently, using the disk names
defined in the configuration file as share points. You must also specify
in the `share` option of the configuration file the Windows UNC path that
remote clients need to use to access these shared disks.

For example, operating from a server named `darkstar`, you can use
the options:

    data d1 F:\array\
    data d2 G:\array\
    data d3 H:\array\
    pool C:\pool
    share \\darkstar

and share the following directories over the network:

    \\darkstar\pool -> C:\pool
    \\darkstar\d1 -> F:\array
    \\darkstar\d2 -> G:\array
    \\darkstar\d3 -> H:\array

to allow remote clients to access all the files at \\darkstar\pool.

You may also need to configure remote clients to enable access to remote
symlinks with the command:

    fsutil behavior set SymlinkEvaluation L2L:1 R2R:1 L2R:1 R2L:1

4.3 Undeleting
--------------

SnapRAID functions more like a backup program than a RAID system, and it
can be used to restore or undelete files to their previous state using
the -f, --filter option:

    snapraid fix -f FILE

or for a directory:

    snapraid fix -f DIR/

You can also use it to recover only accidentally deleted files inside
a directory using the -m, --filter-missing option, which restores
only missing files, leaving all others untouched.

    snapraid fix -m -f DIR/

Or to recover all the deleted files on all drives with:

    snapraid fix -m

4.4 Recovering
--------------

The worst has happened, and you have lost one or more disks!

DO NOT PANIC! You will be able to recover them!

The first thing you must do is avoid further changes to your disk array.
Disable any remote connections to it and any scheduled processes, including
any scheduled SnapRAID nightly sync or scrub.

Then proceed with the following steps.

---- 4.4.1 STEP 1 -> Reconfigure ----
You need some space to recover, ideally on additional
spare disks, but an external USB disk or remote disk will suffice.

Modify the SnapRAID configuration file to make the `data` or `parity`
option of the failed disk point to a location with enough empty
space to recover the files.

For example, if disk `d1` has failed, change from:

    data d1 /mnt/disk1/

to:

    data d1 /mnt/new_spare_disk/

If the disk to recover is a parity disk, update the appropriate `parity`
option.
If you have multiple failed disks, update all their configuration options.

---- 4.4.2 STEP 2 -> Fix ----
Run the fix command, storing the log in an external file with:

    snapraid -d NAME -l fix.log fix

Where NAME is the name of the disk, such as `d1` in our previous example.
If the disk to recover is a parity disk, use the names `parity`, `2-parity`,
etc.
If you have multiple failed disks, use multiple -d options to specify all
of them.

This command will take a long time.

Ensure you have a few gigabytes free to store the fix.log file.
Run it from a disk with sufficient free space.

Now you have recovered all that is recoverable. If some files are partially
or totally unrecoverable, they will be renamed by adding the `.unrecoverable`
extension.

You can find a detailed list of all unrecoverable blocks in the fix.log file
by checking all lines starting with `unrecoverable:`.

If you are not satisfied with the recovery, you can retry it as many
times as you wish.

For example, if you have removed files from the array after the last
`sync`, this may result in some files not being recovered.
In this case, you can retry the `fix` using the -i, --import option,
specifying where these files are now to include them again in the
recovery process.

If you are satisfied with the recovery, you can proceed further,
but note that after syncing, you cannot retry the `fix` command
anymore!

---- 4.4.3 STEP 3 -> Check ----
As a cautious check, you can now run a `check` command to ensure that
everything is correct on the recovered disk.

    snapraid -d NAME -a check

Where NAME is the name of the disk, such as `d1` in our previous example.

The -d and -a options tell SnapRAID to check only the specified disk
and ignore all parity data.

This command will take a long time, but if you are not overly cautious,
you can skip it.

---- 4.4.4 STEP 4 -> Sync ----
Run the `sync` command to resynchronize the array with the new disk.

    snapraid sync

If everything is recovered, this command is immediate.


5 COMMANDS
==========

SnapRAID provides a few simple commands that allow you to:

* Print the status of the array -> `status`
* Control the disks -> `smart`, `probe`, `up`, `down`
* Make a backup/recovery point -> `sync`
* Periodically check data -> `scrub`
* Restore the last backup/recovery point -> `fix`.

Commands must be written in lowercase.

5.1 status
----------

Prints a summary of the state of the disk array.

It includes information about parity fragmentation, how old
the blocks are without checking, and all recorded silent
errors encountered while scrubbing.

The information presented refers to the latest time you
ran `sync`. Later modifications are not taken into account.

If bad blocks were detected, their block numbers are listed.
To fix them, you can use the `fix -e` command.

It also shows a graph representing the last time each block
was scrubbed or synced. Scrubbed blocks are shown with `*`,
blocks synced but not yet scrubbed with `o`.

Nothing is modified.

5.2 smart
---------

Prints a SMART report of all the disks in the system.

It includes an estimation of the probability of failure in the next
year, allowing you to plan maintenance replacements of disks that show
suspicious attributes.

This probability estimation is obtained by correlating the SMART attributes
of the disks with the Backblaze data available at:

    https://www.backblaze.com/hard-drive-test-data.html

If SMART reports that a disk is failing, `FAIL` or `PREFAIL` is printed
for that disk, and SnapRAID returns with an error.
In this case, immediate replacement of the disk is highly recommended.

Other possible status strings are:
    logfail - In the past, some attributes were lower than
        the threshold.
    logerr - The device error log contains errors.
    selferr - The device self-test log contains errors.

If the -v, --verbose option is specified, a deeper statistical analysis
is provided. This analysis can help you decide if you need more
or less parity.

This command uses the `smartctl` tool and is equivalent to running
`smartctl -a` on all devices.

If your devices are not auto-detected correctly, you can specify
a custom command using the `smartctl` option in the configuration
file.

Nothing is modified.

5.3 probe
---------

Prints the POWER state of all disks in the system.

`Standby` means the disk is not spinning. `Active` means
the disk is spinning.

This command uses the `smartctl` tool and is equivalent to running
`smartctl -n standby -i` on all devices.

If your devices are not auto-detected correctly, you can specify
a custom command using the `smartctl` option in the configuration
file.

Nothing is modified.

5.4 up
------

Spins up all the disks of the array.

You can spin up only specific disks using the -d, --filter-disk option.

Spinning up all the disks at the same time requires a lot of power.
Ensure that your power supply can sustain it.

Nothing is modified.

5.5 down
--------

Spins down all the disks of the array.

This command uses the `smartctl` tool and is equivalent to running
`smartctl -s standby,now` on all devices.

You can spin down only specific disks using the -d, --filter-disk
option.

To automatically spin down on error, you can use the -s, --spin-down-on-error
option with any other command, which is equivalent to running `down` manually
when an error occurs.

Nothing is modified.

5.6 diff
--------

Lists all the files modified since the last `sync` that need to have
their parity data recomputed.

This command doesn't check the file data, but only the file timestamp,
size, and inode.

After listing all changed files, a summary of the changes is
presented, grouped by:
    equal - Files unchanged from before.
    added - Files added that were not present before.
    removed - Files removed.
    updated - Files with a different size or timestamp, meaning they
        were modified.
    moved - Files moved or renamed on the same disk while retaining
        the same inode. They are identified by having the same inode,
        size, and timestamp, but a different path. The file name may
        also be different.
    copied - Files copied on the same or a different disk where the
        original file still exists. They are identified by having
        the same name, size, and timestamp. If the sub-second
        timestamp is zero, the full path must match to be
        identified, not just the name.
    relocated - Files moved to a different directory or disk where the
        original file has disappeared and the inode is not retained.
        They are identified by having the same name, size, and timestamp.
        If the sub-second timestamp is zero, the full path must match to
        be identified. Unlike 'moved' files, relocated files require the
        file name to remain the same.
    restored - Files with a different inode but matching directory,
        name, size, and timestamp.
        These are usually files restored after being deleted.

If a `sync` is required, the process return code is 2, instead of the
default 0. The return code 1 is used for a generic error condition.

Nothing is modified.

5.7 sync
--------

Updates the parity information. All modified files
in the disk array are read, and the corresponding parity
data is updated.

You can stop this process at any time by pressing Ctrl+C,
without losing the work already done.
At the next run, the `sync` process will resume where
it was interrupted.

If silent or input/output errors are found during the process,
the corresponding blocks are marked as bad.

Files are identified by path and/or inode and checked by
size and timestamp.
If the file size or timestamp differs, the parity data
is recomputed for the entire file.
If the file is moved or renamed on the same disk, keeping the
same inode, the parity is not recomputed.
If the file is moved to another disk, the parity is recomputed,
but the previously computed hash information is retained.

The `content` and `parity` files are modified if necessary.
The files in the array are NOT modified.

5.8 scrub
---------

Scrubs the array, checking for silent or input/output errors in data
and parity disks.

Each invocation checks approximately 8% of the array, excluding
data already scrubbed in the last 10 days.
This means that scrubbing once a week ensures every bit of data is checked
at least once every three months.

You can define a different scrub plan or amount using the -p, --plan
option, which accepts:
bad - Scrub blocks marked bad.
new - Scrub just-synced blocks not yet scrubbed.
full - Scrub everything.
0-100 - Scrub the specified percentage of blocks.

If you specify a percentage amount, you can also use the -o, --older-than
option to define how old the block should be.
The oldest blocks are scrubbed first, ensuring an optimal check.
If you want to scrub only the just-synced blocks not yet scrubbed,
use the `-p new` option.

To get details of the scrub status, use the `status` command.

For any silent or input/output error found, the corresponding blocks
are marked as bad in the `content` file.
These bad blocks are listed in `status` and can be fixed with `fix -e`.
After the fix, at the next scrub, they will be rechecked, and if found
corrected, the bad mark will be removed.
To scrub only the bad blocks, you can use the `scrub -p bad` command.

It's recommended to run `scrub` only on a synced array to avoid
reported errors caused by unsynced data. These errors are recognized
as not being silent errors, and the blocks are not marked as bad,
but such errors are reported in the output of the command.

The `content` file is modified to update the time of the last check
for each block and to mark bad blocks.
The `parity` files are NOT modified.
The files in the array are NOT modified.

5.9 fix
-------

Fixes all the files and the parity data.

All files and parity data are compared with the state saved in
the last `sync`.
If a difference is found, it is reverted to the stored state.

WARNING! The `fix` command does not differentiate between errors and
intentional modifications. It unconditionally reverts the file state
to the last `sync`.

If no other option is specified, the entire array is processed.
Use the filter options to select a subset of files or disks to operate on.

To fix only the blocks marked bad during `sync` and `scrub`,
use the -e, --filter-error option.
Unlike other filter options, this one applies fixes only to files that are
unchanged since the latest `sync`.

SnapRAID renames all files that cannot be fixed by adding the
`.unrecoverable` extension.

Before fixing, the entire array is scanned to find any files moved
since the last `sync` operation.
These files are identified by their timestamp, ignoring their name
and directory, and are used in the recovery process if necessary.
If you moved some of them outside the array, you can use the -i, --import
option to specify additional directories to scan.

Files are identified only by path, not by inode.

The `content` file is NOT modified.
The `parity` files are modified if necessary.
The files in the array are modified if necessary.

5.10 check
----------

Verifies all the files and the parity data.

It works like `fix`, but it only simulates a recovery and no changes
are written to the array.

This command is primarily intended for manual verification,
such as after a recovery process or in other special conditions.
For periodic and scheduled checks, use `scrub`.

If you use the -a, --audit-only option, only the file
data is checked, and the parity data is ignored for a
faster run.

Files are identified only by path, not by inode.

Nothing is modified.

5.11 list
---------

Lists all the files contained in the array at the time of the
last `sync`.

With -v or --verbose, the subsecond time is also shown.

Nothing is modified.

5.12 dup
--------

Lists all duplicate files. Two files are assumed equal if their
hashes match. The file data is not read; only the
precomputed hashes are used.

Nothing is modified.

5.13 pool
---------

Creates or updates a virtual view of all
the files in your disk array in the `pooling` directory.

The files are not copied but linked using
symbolic links.

When updating, all existing symbolic links and empty
subdirectories are deleted and replaced with the new
view of the array. Any other regular files are left in place.

Nothing is modified outside the pool directory.

5.14 devices
------------

Prints the low-level devices used by the array.

This command displays the device associations in the array
and is mainly intended as a script interface.

The first two columns are the low-level device ID and path.
The next two columns are the high-level device ID and path.
The last column is the disk name in the array.

In most cases, you have one low-level device for each disk in the
array, but in some more complex configurations, you may have multiple
low-level devices used by a single disk in the array.

Nothing is modified.

5.15 touch
----------

Sets an arbitrary sub-second timestamp for all files
that have it set to zero.

This improves SnapRAID's ability to recognize moved
and copied files, as it makes the timestamp almost unique,
reducing possible duplicates.

More specifically, if the sub-second timestamp is not zero,
a moved or copied file is identified as such if it matches
the name, size, and timestamp. If the sub-second timestamp
is zero, it is considered a copy only if the full path,
size, and timestamp all match.

The second-precision timestamp is not modified,
so all the dates and times of your files will be preserved.

5.16 rehash
-----------

Schedules a rehash of the entire array.

This command changes the hash kind used, typically when upgrading
from a 32-bit system to a 64-bit one, to switch from
MurmurHash3 to the faster SpookyHash.

If you are already using the optimal hash, this command
does nothing and informs you that no action is needed.

The rehash is not performed immediately but takes place
progressively during `sync` and `scrub`.

You can check the rehash state using `status`.

During the rehash, SnapRAID maintains full functionality,
with the only exception that `dup` cannot detect duplicated
files using a different hash.

5.17 locate
-----------

Locate files stored in the parity disks. For each matching file, it
prints its location within the parity file and the number of fragments
it occupies.

You can use the -t, --tail option to restrict the operation to files
occupying the specified tail portion of the parity.

If you want to reallocate these files, you can then use the
-W, --force-realloc-tail option. Be aware that such files will
not be protected by parity during the reallocation process.


6 OPTIONS
=========

SnapRAID provides the following options:

    -c, --conf CONFIG
        Selects the configuration file to use. If not specified, in Unix
        it uses the file `/usr/local/etc/snapraid.conf` if it exists,
        otherwise `/etc/snapraid.conf`.
        In Windows, it uses the file `snapraid.conf` in the same
        directory as `snapraid.exe`.

    -f, --filter PATTERN
        Filters the files to process in `check` and `fix`.
        Only the files matching the specified pattern are processed.
        This option can be used multiple times.
        See the PATTERN section for more details on
        pattern specifications.
        In Unix, ensure globbing characters are quoted if used.
        This option can be used only with `check` and `fix`.
        It cannot be used with `sync` and `scrub`, as they always
        process the entire array.

    -d, --filter-disk NAME
        Filters the disks to process in `check`, `fix`, `up`, and `down`.
        You must specify a disk name as defined in the configuration
        file.
        You can also specify parity disks with the names: `parity`, `2-parity`,
        `3-parity`, etc., to limit operations to a specific parity disk.
        If you combine multiple --filter, --filter-disk, and --filter-missing options,
        only files matching all the filters are selected.
        This option can be used multiple times.
        This option can be used only with `check`, `fix`, `up`, and `down`.
        It cannot be used with `sync` and `scrub`, as they always
        process the entire array.

    -m, --filter-missing
        Filters the files to process in `check` and `fix`.
        Only the files missing or deleted from the array are processed.
        When used with `fix`, this acts as an `undelete` command.
        If you combine multiple --filter, --filter-disk, and --filter-missing options,
        only files matching all the filters are selected.
        This option can be used only with `check` and `fix`.
        It cannot be used with `sync` and `scrub`, as they always
        process the entire array.

    -e, --filter-error
        Processes the files with errors in `check` and `fix`.
        It processes only files that have blocks marked with silent
        or input/output errors during `sync` and `scrub`, as listed in `status`.
        This option can be used only with `check` and `fix`.

    -p, --plan PERC|bad|new|full
        Selects the scrub plan. If PERC is a numeric value from 0 to 100,
        it is interpreted as the percentage of blocks to scrub.
        Instead of a percentage, you can specify a plan:
        `bad` scrubs bad blocks, `new` scrubs blocks not yet scrubbed,
        and `full` scrubs everything.
        This option can be used only with `scrub`.

    -o, --older-than DAYS
        Selects the oldest part of the array to process in `scrub`.
        DAYS is the minimum age in days for a block to be scrubbed;
        the default is 10.
        Blocks marked as bad are always scrubbed regardless of this option.
        This option can be used only with `scrub`.

    -a, --audit-only
        In `check`, verifies the hash of the files without
        checking the parity data.
        If you are interested only in checking the file data, this
        option can significantly speed up the checking process.
        This option can be used only with `check`.

    -h, --pre-hash
        In `sync`, runs a preliminary hashing phase of all new data
        for additional verification before the parity computation.
        Usually, in `sync`, no preliminary hashing is done, and the new
        data is hashed just before the parity computation when it is read
        for the first time.
        This process occurs when the system is under
        heavy load, with all disks spinning and a busy CPU.
        This is an extreme condition for the machine, and if it has a
        latent hardware problem, silent errors may go undetected
        because the data is not yet hashed.
        To avoid this risk, you can enable the `pre-hash` mode to have
        all the data read twice to ensure its integrity.
        This option also verifies files moved within the array
        to ensure the move operation was successful and, if necessary,
        allows you to run a fix operation before proceeding.
        This option can be used only with `sync`.

    -i, --import DIR
        Imports from the specified directory any files deleted
        from the array after the last `sync`.
        If you still have such files, they can be used by `check`
        and `fix` to improve the recovery process.
        The files are read, including in subdirectories, and are
        identified regardless of their name.
        This option can be used only with `check` and `fix`.

    -s, --spin-down-on-error
        On any error, spins down all managed disks before exiting with
        a non-zero status code. This prevents the drives from
        remaining active and spinning after an aborted operation,
        helping to avoid unnecessary heat buildup and power
        consumption. Use this option to ensure disks are safely
        stopped even when a command fails.

    -w, --bw-limit RATE
        Applies a global bandwidth limit for all disks. The RATE is
        the number of bytes per second. You can specify a multiplier
        such as K, M, G, or T (e.g., --bw-limit 1G).

    -t, --tail SIZE
        Limit file listing to those using no more than the specified
        tail size of the parity disks.
        You can use multipliers such as K, M, G, or T (e.g. --tail 1G).
        This option is only valid when used together with the `locate`
        command.

    -A, --stats
        Enables an extended status view that shows additional information.
        The screen displays two graphs:
        The first graph shows the number of buffered stripes for each
        disk, along with the file path of the file currently being
        accessed on that disk. Typically, the slowest disk will have
        no buffer available, which determines the maximum achievable
        bandwidth.
        The second graph shows the percentage of time spent waiting
        over the past 100 seconds. The slowest disk is expected to
        cause most of the wait time, while other disks should have
        little or no wait time because they can use their buffered stripes.
        This graph also shows the time spent waiting for hash
        calculations and RAID computations.
        All computations run in parallel with disk operations.
        Therefore, as long as there is measurable wait time for at
        least one disk, it indicates that the CPU is fast enough to
        keep up with the workload.

    -Z, --force-zero
        Forces the insecure operation of syncing a file with zero
        size that was previously non-zero.
        If SnapRAID detects such a condition, it stops proceeding
        unless you specify this option.
        This allows you to easily detect when, after a system crash,
        some accessed files were truncated.
        This is a possible condition in Linux with the ext3/ext4
        file systems.
        This option can be used only with `sync`.

    -E, --force-empty
        Forces the insecure operation of syncing a disk with all
        the original files missing.
        If SnapRAID detects that all the files originally present
        on the disk are missing or rewritten, it stops proceeding
        unless you specify this option.
        This allows you to easily detect when a data file system is not
        mounted.
        This option can be used only with `sync`.

    -U, --force-uuid
        Forces the insecure operation of syncing, checking, and fixing
        with disks that have changed their UUID.
        If SnapRAID detects that some disks have changed UUID,
        it stops proceeding unless you specify this option.
        This allows you to detect when your disks are mounted at the
        wrong mount points.
        It is, however, allowed to have a single UUID change with
        single parity, and more with multiple parity, because this is
        the normal case when replacing disks after a recovery.
        This option can be used only with `sync`, `check`, or
        `fix`.

    -D, --force-device
        Forces the insecure operation of fixing with inaccessible disks
        or with disks on the same physical device.
        For example, if you lost two data disks and have a spare disk to recover
        only the first one, you can ignore the second inaccessible disk.
        Or, if you want to recover a disk in the free space left on an
        already used disk, sharing the same physical device.
        This option can be used only with `fix`.

    -N, --force-nocopy
        In `sync`, `check`, and `fix`, disables the copy detection heuristic.
        Without this option, SnapRAID assumes that files with the same
        attributes, such as name, size, and timestamp, are copies with the
        same data.
        This allows identification of copied or moved files from one disk
        to another and reuses the already computed hash information
        to detect silent errors or to recover missing files.
        In some rare cases, this behavior may result in false positives
        or a slow process due to many hash verifications, and this
        option allows you to resolve such issues.
        This option can be used only with `sync`, `check`, and `fix`.

    -F, --force-full
        In `sync`, forces a full recomputation of the parity.
        This option can be used when you add a new parity level or if
        you reverted to an old content file using more recent parity data.
        Instead of recreating the parity from scratch, this allows
        you to reuse the hashes present in the content file to validate data
        and maintain data protection during the `sync` process using
        the existing parity data.
        This option can be used only with `sync`.

    -R, --force-realloc
        In `sync`, forces a full reallocation of files and rebuild of the parity.
        This option can be used to completely reallocate all files,
        removing fragmentation, while reusing the hashes present in the content
        file to validate data.
        This option can be used only with `sync`.
        WARNING! This option is for experts only, and it is highly
        recommended not to use it.
        You DO NOT have data protection during the `sync` operation.

    -W, --force-realloc-tail SIZE
        Works like -R, --force-realloc, but limited to the specified
        tail portion (last SIZE bytes) of each parity file.
        It forces reallocation (movement) of any file fragments/blocks
        currently stored in that tail section, allowing them to be
        placed anywhere in the parity file(s) where free space is
        available (including existing holes).
        The main purpose of this option is to shrink the on-disk size
        of the parity file. If the reallocation successfully clears
        the entire tail section (no blocks remain using it), the
        parity file is truncated, reclaiming the unused tail space.
        You can use multipliers such as K, M, G, or T (e.g.
        --force-realloc-tail 1G).
        You can use locate -t, --tail to know in advance the affected
        files.
        WARNING! This option is for experts only, and it is highly
        recommended not to use it.
        You DO NOT have data protection during the `sync` operation
        for the affected files.

    -l, --log FILE
        Writes a detailed log to the specified file.
        If this option is not specified, unexpected errors are printed
        to the screen, potentially resulting in excessive output in case of
        many errors. When -l, --log is specified, only
        fatal errors that cause SnapRAID to stop are printed
        to the screen.
        If the path starts with `>>`, the file is opened
        in append mode. Occurrences of `%D` and `%T` in the name are
        replaced with the date and time in the format YYYYMMDD and
        HHMMSS. In Windows batch files, you must double
        the `%` character, e.g., result-%%D.log. To use `>>`, you must
        enclose the name in quotes, e.g., `">>result.log"`.
        To output the log to standard output or standard error,
        you can use `">&1"` and `">&2"`, respectively.
        See the snapraid_log.txt file or man page for log tag descriptions.

    -L, --error-limit NUMBER
        Sets a new error limit before stopping execution.
        By default, SnapRAID stops if it encounters more than 100
        input/output errors, indicating that a disk is likely failing.
        This option affects `sync` and `scrub`, which are allowed
        to continue after the first set of disk errors to try
        to complete their operations.
        However, `check` and `fix` always stop at the first error.

    -S, --start BLKSTART
        Starts processing from the specified
        block number. This can be useful for retrying to check
        or fix specific blocks in case of a damaged disk.
        This option is mainly for advanced manual recovery and
        is supported only by `check` and `fix`.

    -B, --count BLKCOUNT
        Processes only the specified number of blocks.
        This option is mainly for advanced manual recovery and
        is supported only by `check` and `fix`.

    -C, --gen-conf CONTENT
        Generates a dummy configuration file from an existing
        content file.
        The configuration file is written to standard output
        and does not overwrite an existing one.
        This configuration file also contains the information
        needed to reconstruct the disk mount points in case you
        lose the entire system.

    -v, --verbose
        Prints more information to the screen.
        If specified once, it prints excluded files
        and additional statistics.
        This option has no effect on the log files.

    -q, --quiet
        Prints less information to the screen.
        If specified once, it removes the progress bar; twice,
        the running operations; three times, the info
        messages; four times, the status messages.
        Fatal errors are always printed to the screen.
        This option has no effect on the log files.

    -H, --help
        Prints a short help screen.

    -V, --version
        Prints the program version.


7 CONFIGURATION
===============

SnapRAID requires a configuration file to know where your disk array
is located and where to store the parity information.

In Unix, it uses the file `/usr/local/etc/snapraid.conf` if it exists,
otherwise `/etc/snapraid.conf`.
In Windows, it uses the file `snapraid.conf` in the same
directory as `snapraid.exe`.

It must contain the following options (case-sensitive):

7.1 parity FILE [,FILE] ...
---------------------------

Defines the files to use to store the parity information.
The parity enables protection from a single disk
failure, similar to RAID5.

You can specify multiple files, which must be on different disks.
When a file cannot grow anymore, the next one is used.
The total space available must be at least as large as the largest data disk in
the array.

You can add additional parity files later, but you
cannot reorder or remove them.

Keeping the parity disks reserved for parity ensures that
they do not become fragmented, improving performance.

In Windows, 256 MB is left unused on each disk to avoid the
warning about full disks.

This option is mandatory and can be used only once.

7.2 (2,3,4,5,6)-parity FILE [,FILE] ...
---------------------------------------

Defines the files to use to store extra parity information.

For each parity level specified, one additional level of protection
is enabled:

* 2-parity enables RAID6 dual parity.
* 3-parity enables triple parity.
* 4-parity enables quad (four) parity.
* 5-parity enables penta (five) parity.
* 6-parity enables hexa (six) parity.

Each parity level requires the presence of all previous parity
levels.

The same considerations as for the `parity` option apply.

These options are optional and can be used only once.

7.3 z-parity FILE [,FILE] ...
-----------------------------

Defines an alternate file and format to store triple parity.

This option is an alternative to `3-parity`, primarily intended for
low-end CPUs like ARM or AMD Phenom, Athlon, and Opteron that do not
support the SSSE3 instruction set. In such cases, it provides
better performance.

This format is similar to but faster than the one used by ZFS RAIDZ3.
Like ZFS, it does not work beyond triple parity.

When using `3-parity`, you will be warned if it is recommended to use
the `z-parity` format for performance improvement.

It is possible to convert from one format to another by adjusting
the configuration file with the desired z-parity or 3-parity file
and using `fix` to recreate it.

7.4 content FILE
----------------

Defines the file to use to store the list and checksums of all the
files present in your disk array.

It can be placed on a disk used for data, parity, or
any other disk available.
If you use a data disk, this file is automatically excluded
from the `sync` process.

This option is mandatory and can be used multiple times to save
multiple copies of the same file.

You must store at least one copy for each parity disk used
plus one. Using additional copies does not hurt.

7.5 data NAME DIR
-----------------

Defines the name and mount point of the data disks in
the array. NAME is used to identify the disk and must
be unique. DIR is the mount point of the disk in the
file system.

You can change the mount point as needed, as long as
you keep the NAME fixed.

However, the configured path must consistently resolve to the
same underlying data. If the path contains symbolic links,
their targets must remain stable. Retargeting such symlinks
to switch the data directory on the fly is unsupported.

You should use one option for each data disk in the array.

You can rename a disk later by changing the NAME directly
in the configuration file and then running a `sync` command.
In the case of renaming, the association is done using the stored
UUID of the disks.

7.6 extra NAME DIR
------------------

Defines the name and mount point of additional disks to monitor
with the `smart` and `probe` commands.

This is useful for monitoring disks that are not part of the
array but are required for the system to function, such as
the boot disk.

Note that such disks are not affected by the `up` and `down` commands
because they are expected to be always spinning.

7.7 snapshot
------------

Enable the use of filesystem snapshots for the `sync`, `scrub`,
`check`, and `fix` commands.

When enabled, SnapRAID creates a read-only snapshot of each supported
data filesystem at the start of a 'sync'. This provides a stable view of
each filesystem, preventing later modifications from changing the data
used for parity computation.

This significantly improves recovery: if a file is deleted from the live
filesystem, it remains preserved in the snapshot. This prevents the parity
from becoming "broken" for that block, ensuring you can still successfully
recover data if another disk fails.

This option applies exclusively to data disks formatted with the
Btrfs, Bcachefs or ZFS filesystems in Linux and NTFS in Windows.
Parity disks, or data disks using other filesystems, will always use
the live version of the filesystem.

Snapshot creation and deletion require administrative privileges.
Ensure SnapRAID is run with the necessary permissions (e.g., sudo) when
this option is enabled.

See the SNAPSHOTS section for a detailed explanation of the snapshot
lifecycle.

7.8 nohidden
------------

Excludes all hidden files and directories.
In Unix, hidden files are those starting with `.`.
In Windows, they are those with the hidden attribute.

7.9 exclude/include PATTERN
---------------------------

Defines the file or directory patterns to exclude or include
in the sync process.
All patterns are processed in the specified order.

If the first pattern that matches is an `exclude` one, the file
is excluded. If it is an `include` one, the file is included.
If no pattern matches, the file is excluded if the last pattern
specified is an `include`, or included if the last pattern
specified is an `exclude`.

Note that filtering occurs hierarchically during directory traversal.
If a directory matches an `exclude` rule, the entire directory is
skipped (pruned), and any `include` rules for files or subdirectories
inside it will not be evaluated.

See the PATTERN section for more details on pattern
specifications.

This option can be used multiple times.

7.10 blocksize SIZE_IN_KIBIBYTES
--------------------------------

Defines the basic block size in kibibytes for the parity.
One kibibyte is 1024 bytes.

The default blocksize is 256, which should work for most cases.

WARNING! This option is for experts only, and it is highly
recommended not to change this value. To change this value in the
future, you will need to recreate the entire parity!

A reason to use a different blocksize is if you have many small
files, on the order of millions.

For each file, even if only a few bytes, an entire block of parity is allocated,
and with many files, this may result in significant unused parity space.
When you completely fill the parity disk, you are not
allowed to add more files to the data disks.
However, the wasted parity does not accumulate across data disks. Wasted space
resulting from a high number of files on a data disk limits only
the amount of data on that data disk, not others.

As an approximation, you can assume that half of the block size is
wasted for each file. For example, with 100,000 files and a 256 KiB
block size, you will waste 12.8 GB of parity, which may result
in 12.8 GB less space available on the data disk.

You can check the amount of wasted space on each disk using `status`.
This is the amount of space you must leave free on the data
disks or use for files not included in the array.
If this value is negative, it means you are close to filling
the parity, and it represents the space you can still waste.

To avoid this issue, you can use a larger partition for parity.
For example, if the parity partition is 12.8 GB larger than the data disks,
you have enough extra space to handle up to 100,000
files on each data disk without any wasted space.

A trick to get a larger parity partition in Linux is to format it
with the command:

    mkfs.ext4 -m 0 -T largefile4 DEVICE

This results in about 1.5% extra space, approximately 60 GB for
a 4 TB disk, which allows about 460,000 files on each data disk without
any wasted space.

7.11 hashsize SIZE_IN_BYTES
---------------------------

Defines the hash size in bytes for the saved blocks.

The default hashsize is 16 bytes (128 bits), which should work
for most cases.

WARNING! This option is for experts only, and it is highly
recommended not to change this value. To change this value in the
future, you will need to recreate the entire parity!

A reason to use a different hashsize is if your system has
limited memory. As a rule of thumb, SnapRAID typically requires
1 GiB of RAM for each 16 TB of data in the array.

Specifically, to store the hashes of the data, SnapRAID requires
approximately TS*(1+HS)/BS bytes of RAM,
where TS is the total size in bytes of your disk array, BS is the
block size in bytes, and HS is the hash size in bytes.

For example, with 8 disks of 4 TB, a block size of 256 KiB
(1 KiB = 1024 bytes), and a hash size of 16, you get:

RAM = (8 * 4 * 10^12) * (1+16) / (256 * 2^10) = 1.93 GiB

Switching to a hash size of 8, you get:

RAM = (8 * 4 * 10^12) * (1+8) / (256 * 2^10) = 1.02 GiB

Switching to a block size of 512, you get:

RAM = (8 * 4 * 10^12) * (1+16) / (512 * 2^10) = 0.96 GiB

Switching to both a hash size of 8 and a block size of 512, you get:

RAM = (8 * 4 * 10^12) * (1+8) / (512 * 2^10) = 0.51 GiB

7.12 autosave SIZE_IN_GIGABYTES
-------------------------------

Automatically saves the state when syncing or scrubbing after the
specified amount of GB processed.
This option is useful to avoid restarting long `sync`
commands from scratch if interrupted by a machine crash or any other event.

7.13 temp_limit TEMPERATURE_CELSIUS
-----------------------------------

Sets the maximum allowed disk temperature in Celsius. When specified,
SnapRAID periodically checks the temperature of all disks using the
smartctl tool. The current disk temperatures are displayed while
SnapRAID is operating. If any disk exceeds this limit, all operations
stop, and the disks are spun down (put into standby) for the duration
defined by the `temp_sleep` option. After the sleep period, operations
resume, potentially pausing again if the temperature limit is reached
once more.

During operation, SnapRAID also analyzes the heating curve of each
disk and estimates the long-term steady temperature they are expected
to reach if activity continues. The estimation is performed only after
the disk temperature has increased four times, ensuring that enough
data points are available to establish a reliable trend.
This predicted steady temperature is shown in parentheses next to the
current value and helps assess whether the system's cooling is
adequate. This estimated temperature is for informational purposes
only and has no effect on the behavior of SnapRAID. The program's
actions are based solely on the actual measured disk temperatures.

To perform this analysis, SnapRAID needs a reference for the system
temperature. It first attempts to read it from available hardware
sensors. If no system sensor can be accessed, it uses the lowest disk
temperature measured at the start of the run as a fallback reference.

Normally, SnapRAID shows only the temperature of the hottest disk.
To display the temperature of all disks, use the -A or --stats option.

7.14 temp_sleep TIME_IN_MINUTES
-------------------------------

Sets the standby time, in minutes, when the temperature limit is
reached. During this period, the disks remain spun down. The default
is 5 minutes.

7.15 pool DIR
-------------

Defines the pooling directory where the virtual view of the disk
array is created using the `pool` command.

The directory must already exist.

7.16 share UNC_DIR
------------------

Defines the Windows UNC path required to access the disks remotely.

If this option is specified, the symbolic links created in the pool
directory use this UNC path to access the disks.
Without this option, the symbolic links generated use only local paths,
which does not allow sharing the pool directory over the network.

The symbolic links are formed using the specified UNC path, adding the
disk name as specified in the `data` option, and finally adding the
file directory and name.

This option is required only for Windows.

7.17 smartctl DISK/PARITY OPTIONS...
------------------------------------

Defines custom smartctl options to obtain the SMART attributes for
each disk. This may be required for RAID controllers and some USB
disks that cannot be auto-detected. The %s placeholder is replaced by
the device name, but it is optional for fixed devices like RAID controllers.

DISK is the same disk name specified in the `data` option.
PARITY is one of the parity names: `parity`, `2-parity`, `3-parity`,
`4-parity`, `5-parity`, `6-parity`, or `z-parity`.

In the specified OPTIONS, the `%s` string is replaced by the
device name. For RAID controllers, the device is
likely fixed, and you may not need to use `%s`.

You can also specify a custom info options string using the
`[info: ...]` tag to override the default `-a` option.

Refer to the smartmontools documentation for possible options:

    https://www.smartmontools.org/wiki/Supported_RAID-Controllers
    https://www.smartmontools.org/wiki/Supported_USB-Devices

For example:

    smartctl d1 [info: -H -i -c -A] -d sat %s
    smartctl parity -d sat %s

7.18 smartignore DISK/PARITY ATTR [ATTR...]
-------------------------------------------

Ignores the specified SMART attribute when computing the probability
of disk failure. This option is useful if a disk reports unusual or
misleading values for a particular attribute.

DISK is the same disk name specified in the `data` option.
PARITY is one of the parity names: `parity`, `2-parity`, `3-parity`,
`4-parity`, `5-parity`, `6-parity`, or `z-parity`.
The special value * can be used to ignore the attribute on all disks.

Multiple attributes can be specified separated by spaces.
Each attribute can be specified as a number from 1 to 255, or as a
case-insensitive name (e.g. `Current_Pending_Sector`).

For example, to ignore the `Current Pending Sector Count` and
`Reallocated Sectors Count` attributes on all disks by their number
or name:

    smartignore * 197 5
    smartignore * Current_Pending_Sector Reallocated_Sector_Ct

To ignore them only on the first parity disk:

    smartignore parity 197 5

7.19 Examples
-------------

An example of a typical configuration for Unix is:

    parity /mnt/diskp/snapraid.parity
    content /mnt/diskp/snapraid.content
    content /var/snapraid/snapraid.content
    data d1 /mnt/disk1/
    data d2 /mnt/disk2/
    data d3 /mnt/disk3/
    exclude /lost+found/
    exclude /tmp/
    smartctl d1 [info: -H -i -c -A] -d sat %s
    smartctl d2 -d usbjmicron %s
    smartctl parity -d areca,1/1 /dev/sg0
    smartctl 2-parity -d areca,2/1 /dev/sg0

An example of a typical configuration for Windows is:

    parity E:\snapraid.parity
    content E:\snapraid.content
    content C:\snapraid\snapraid.content
    data d1 G:\array\
    data d2 H:\array\
    data d3 I:\array\
    exclude Thumbs.db
    exclude \$RECYCLE.BIN
    exclude \System Volume Information
    smartctl d1 [info: -H -i -c -A] -d sat %s
    smartctl d2 -d usbjmicron %s
    smartctl parity -d areca,1/1 /dev/arcmsr0
    smartctl 2-parity -d areca,2/1 /dev/arcmsr0


8 SNAPSHOTS
===========

If the snapshot option is enabled in the configuration, SnapRAID
uses filesystem snapshots to keep the data read during an operation
stable, even when files on the live filesystems are subsequently
modified or deleted.

The management of snapshots is completely automatic and transparent.
You can continue to use SnapRAID exactly as before, with the
additional protection provided by snapshots handled entirely in
the background.

This provides two primary benefits:

Consistency - Files modified on the live filesystem during a long-running
    sync or scrub will not cause parity mismatches or aborted
    operations, because each snapshotted filesystem remains unchanged
    for the duration of the operation.
Recovery - If a file is updated or deleted from the live filesystem,
    it remains preserved in the snapshot. If a disk failure occurs
    before the next sync is run, SnapRAID uses the data preserved
    in the snapshot to reconstruct the failed disk.

    In case of a disk failure during an active sync process,
    SnapRAID  is also able to read automatically from both the
    snapshot of the previous parity computation and the current one.
    This maximizes the probability of a full recovery by
    providing access to the exact data blocks required to solve
    the parity equations, even if those blocks were modified
    or deleted between syncs.

    Without snapshots, an updated or deleted file on a healthy
    disk results in missing data blocks that may be required to
    fix other failed disks.

Snapshots are created only for data disks and only if the underlying
filesystem supports this functionality. Parity disks always use
the live filesystem.

At present, this is supported on Btrfs, Bcachefs, and ZFS in Linux
and on NTFS in Windows.

You can mix data disks with different filesystems. Only those that
support snapshots will utilize them, while other disks will continue
to operate directly on the live filesystem.

Snapshot creation and deletion require administrative privileges.
Ensure SnapRAID is run with the necessary permissions (e.g., sudo)
when snapshots are enabled.

8.1 Command Behavior with Snapshots
-----------------------------------

The `sync` and `scrub` commands use snapshots for data disks whose
filesystems support them. On a snapshotted disk, the command sees
a stable filesystem image unaffected by concurrent changes to the
live filesystem.
During a `sync`, a new snapshot is created independently for each
supported data filesystem. These snapshots collectively provide the
data used for parity computation.
In contrast, the `scrub` command utilizes the last snapshot, the one
created during the most recent sync, to maintain a reliable reference
point that matches the existing parity.

For the `check` and `fix` commands, the use of the last snapshot
depends on whether specific disks are targeted using the
-d, --filter-disk option.

A disk explicitly selected via -d (the "target" of the operation)
always uses the live filesystem. For `fix`, this allows restoring
data to the active disk replacement. For `check`, it allows simulating
the `fix` operation under the same conditions.

All other data disks (the "reference" disks) will be accessed via
their snapshots. This ensures that even if you are modifying files
on your healthy disks while a recovery is in progress, SnapRAID
has a stable, frozen reference to solve the parity equations.

If no -d option is provided, SnapRAID assumes the operation applies
to the entire array, in this case, `check` and `fix` will use the
live filesystems exclusively.

All other commands operate exclusively on the live filesystem.


9 SNAPSHOTS LIFECYCLE
=====================

SnapRAID manages three specific snapshots, `stable`, `pending`, and `scan`,
within a hidden directory at the root of each data subvolume.
In Btrfs, Bcachefs, and NTFS it's used the `.snapraid/` directory,
in ZFS the standard `.zfs/snapshot/`.

The `stable` snapshot represents the state of the last
successfully completed `sync`, containing the exact data used to
compute the current parity. It serves as the primary data source
for `scrub`, `check`, and `fix` commands.

The `scan` snapshot is a temporary image created at the start
of a `sync` to provide a frozen state for scanning and parity computation.
After scanning and saving the new content state, it replaces the previous
`pending` snapshot.

The `pending` snapshot is the image used by an interrupted sync.
Upon successful completion of the operation, the previous `stable`
snapshot is deleted, and the `pending` snapshot is promoted to
take its place as the new stable reference.

If a `sync` is interrupted, the `pending` snapshot is preserved.
Subsequent `scrub`, `check`, and `fix` commands will use this
pending snapshot as it matches the filesystem state recorded in
the .content file, even if the parity is only partially synchronized.

In the event of a `sync` interruption, `check` and `fix` commands
also read from the `stable` snapshot to retrieve data blocks
from files that were updated or deleted during the sync.
This is possible because the .content file retains metadata for
all modified or deleted files throughout the sync process.
This historical information is only cleared once a sync finishes
successfully and the final version of the .content file is saved.

If a `sync` is restarted after an interruption, the existing
pending snapshot is deleted and a new one is created to
capture the current state of the live filesystem.
However, if a `sync` is interrupted, avoid modifying, deleting,
replacing, or moving files until a subsequent `sync` completes
successfully. Parity may still contain blocks computed from the
snapshot used by the interrupted sync. A later sync may replace
that snapshot, making intermediate file versions unavailable for
recovery.

9.1 External backups with filesystem snapshots
----------------------------------------------

When filesystem snapshots are enabled, SnapRAID accesses or creates
snapshots at the root of each data subvolume, volume, or dataset.

A simple way to avoid any interference with external tools is to point
your `data` configuration to a regular subdirectory within the subvolume
or volume (for example, `/mnt/subvol/data` instead of `/mnt/subvol`).
SnapRAID always creates the snapshot directory at the subvolume root
(for example, `/mnt/subvol/.snapraid/`), leaving the data subdirectory
clean so that snapshots remain outside and invisible to backup tools
operating inside it.

Otherwise, if your configured `data` directory is itself the subvolume
or volume root, SnapRAID creates the `.snapraid/` directory directly
inside it (on Btrfs, Bcachefs, and NTFS). While SnapRAID automatically
excludes this directory from its own scanning, external backup tools
or file synchronizers will encounter it. If not excluded, they may
traverse into the snapshots or directory links and copy the entire
frozen data tree, duplicating data and potentially exhausting backup
storage. Always configure external backup tools to exclude the
`.snapraid/` directory.

On ZFS, snapshots are exposed through the standard `.zfs/snapshot/`
virtual directory rather than `.snapraid/`. By default (`snapdir=hidden`),
ZFS hides this directory from directory listings, so external backup
tools will not traverse it. If you have enabled `snapdir=visible`,
ensure your backup tools exclude the `.zfs/` directory.


10 PATTERN
==========

Patterns are used to select a subset of files to exclude or include in
the process. Globbing characters can be used to match files and paths
in a flexible way.

The question mark `?` matches any single character except the directory
separator. This makes it useful for matching filenames with variable
characters while keeping the pattern confined to a single directory level.

The single star `*` matches any sequence of characters, but like the
question mark, it never crosses directory boundaries. It stops at the
forward slash, making it suitable for matching within a single path
component. This is the standard wildcard behavior familiar from shell
globbing.

The double star `**` wildcard matches across directory separators only
when it is adjacent to a directory separator (preceded or followed by a
slash). In this form, it matches zero or more characters across directory
levels. For example, `foo/**/bar` matches `foo/bar`, `foo/a/bar`, and
`foo/a/b/bar`; `**/build` matches `build` and `foo/bar/build`; and
`src/**.js` matches `src/main.js` and `src/ui/components/main.js`.
The standalone form `/**/` is not accepted as a complete pattern.

When `**` is not adjacent to a slash, it behaves like `*` and does not
match directory separators. For example, `**.jpg` matches `photo.jpg`
within the current directory, but does not match `dir/photo.jpg`. Use
`**/*.jpg` for that purpose.

Character classes using square brackets match a single character from a
specified set or range. Like the other single character patterns, they do
not match directory separators. Classes support ranges and negation using
an exclamation mark.

The fundamental distinction to remember is that `*`, `?`, and character
classes all respect directory boundaries and only match within a single path
component, while `**` can match across directory separators only when
adjacent to a slash.

There are four different types of patterns:

    FILE
        Selects any file named FILE.
        This pattern applies only to files, not directories.

    DIR/
        Selects any directory named DIR and everything inside.
        This pattern applies only to directories, not files.

    /PATH/FILE
        Selects the exact specified file path. This pattern applies
        only to files, not directories.

    /PATH/DIR/
        Selects the exact specified directory path and everything
        inside.	This pattern applies only to directories, not files.

When you specify an absolute path starting with /, it is applied at
the array root directory, not the local file system root directory.

In Windows, you can use the backslash \ instead of the forward slash /.
Windows system directories, junctions, mount points, and other Windows
special directories are treated as files, meaning that to exclude
them, you must use a file rule, not a directory one.

If the file name contains a `*`, `?`, `[`, or `]` character, you must
escape it to avoid having it interpreted as a globbing character.
In Unix, the escape character is `\`; in Windows, it is `^`.
When the pattern is on the command line, you must double the escape
character to avoid having it interpreted by the command shell.

In the configuration file, you can use different strategies to filter
the files to process.
The simplest approach is to use only `exclude` rules to remove all the
files and directories you do not want to process. For example:

    # Excludes any file named `*.unrecoverable`
    exclude *.unrecoverable
    # Excludes the root directory `/lost+found`
    exclude /lost+found/
    # Excludes any subdirectory named `tmp`
    exclude tmp/

The opposite approach is to define only the files you want to process, using
only `include` rules. For example:

    # Includes only some directories
    include /movies/
    include /musics/
    include /pictures/

The final approach is to mix `exclude` and `include` rules. In this case,
the order of rules is important. Earlier rules take
precedence over later ones.
To simplify, you can list all the `exclude` rules first and then
all the `include` rules. For example:

    # Excludes any file named `*.unrecoverable`
    exclude *.unrecoverable
    # Excludes any subdirectory named `tmp`
    exclude tmp/
    # Includes only some directories
    include /movies/
    include /musics/
    include /pictures/

Note that excluding a directory (using DIR/ or /PATH/DIR/) prevents
SnapRAID from entering and scanning that directory. Consequently, it is
not possible to re-include a file or subdirectory located inside an
excluded parent directory, because the scanner never traverses into it.

For example, the following configuration will NOT include `file.dat`
because `/keep/` is pruned before `file.dat` can be seen:

    # Does NOT work: /keep/ is skipped entirely
    include /keep/file.dat
    exclude /keep/

To exclude files inside a directory while keeping specific ones, match
the files within the directory rather than excluding the directory itself:

    # Works: /keep/ is traversed, but other files inside are excluded
    include /keep/file.dat
    exclude /keep/*

On the command line, using the -f option, you can only use `include`
patterns. For example:

    # Checks only the .mp3 files.
    # In Unix, use quotes to avoid globbing expansion by the shell.
    snapraid -f "*.mp3" check

In Unix, when using globbing characters on the command line, you must
quote them to prevent the shell from expanding them.


11 IGNORE FILE
==============

In addition to the global rules in the configuration file, you can
place `.snapraidignore` files in any directory within the array to
define decentralized exclusion rules.

Rules defined in `.snapraidignore` are applied after the rules in the
configuration file. This means they have a higher priority and can be
used to exclude files that were previously included by the global
configuration. Effectively, if a local rule matches, the file is
excluded regardless of the global include settings.

The pattern logic in `.snapraidignore` mirrors the global configuration
but anchors patterns to the directory where the file is located:

    FILE
        Selects any file named FILE in this directory or below.
        This follows the same globbing rules as the global pattern.

    DIR/
        Selects any directory named DIR and everything inside,
        residing in this directory or below.

    /PATH/FILE
        Selects the exact specified file relative to the location
        of the `.snapraidignore` file.

    /PATH/DIR/
        Selects the exact specified directory and everything inside,
        relative to the location of the `.snapraidignore` file.

Unlike the global configuration, `.snapraidignore` files only support
exclusion rules; you cannot use `include` patterns or negation (!).

For example, if you have a `.snapraidignore` in `/mnt/disk1/projects/`:

    # Excludes ONLY /mnt/disk1/projects/output.bin
    /output.bin
    # Excludes any directory named `build` inside projects/
    build/
    # Excludes any .tmp file inside projects/ or its subfolders
    *.tmp


12 CONTENT
==========

SnapRAID stores the list and checksums of your files in the content file.

It is a binary file that lists all the files present in your disk array,
along with all the checksums to verify their integrity.

The content file is read and written by the `sync`, `scrub`, `touch`, and
`rehash` commands. It is read by the `status`, `diff`, `check`, `fix`,
`list`, `dup`, `locate`, and `pool` commands, and it is completely
ignored by the `smart`, `probe`, `up`, `down`, and `devices` commands.


13 PARITY
=========

SnapRAID stores the parity information of your array in the parity
files.

These are binary files containing the computed parity of all the
blocks defined in the `content` file.

These files are read and written by the `sync` and `fix` commands and
only read by the `scrub` and `check` commands.


14 ENCODING
===========

SnapRAID in Unix ignores any encoding. It reads and stores the
file names with the same encoding used by the file system.

In Windows, all names read from the file system are converted and
processed in UTF-8 format.

To have file names printed correctly, you must set the Windows
console to UTF-8 mode with the command `chcp 65001` and use
a TrueType font like `Lucida Console` as the console font.
This affects only the printed file names; if you
redirect the console output to a file, the resulting file is always
in UTF-8 format.


15 EXIT CODE
============

SnapRAID terminates with the following error codes:

0 - Everything OK.
1 - The command encountered some errors.
2 - The `diff` command found everything is OK, but a `sync` is
    needed.
3 - A disk device command (`probe`, `smart`, `devices`, `up`,
    `down`) completed, but at least one filesystem is running in
    degraded mode with missing device(s).


16 KNOWN ISSUES
===============

16.1 Filesystem boundaries
--------------------------

SnapRAID does not cross filesystem, mount, or subvolume boundaries while
scanning a data disk. When such a boundary is detected, SnapRAID reports
a WARNING and does not scan beyond it.

Files located beyond such boundaries are not included in the data disk and
must be configured separately if they have to be protected.

16.2 Mount points with filesystem snapshots
-------------------------------------------

When filesystem snapshots are enabled, nested mount points and bind mounts
inside a data disk are not supported.

Mount points are part of the live system mount namespace and are not
preserved in filesystem snapshots. Consequently, when scanning a snapshot,
a directory hidden by a mount point in the live filesystem may become
visible.

Do not place nested mount points or bind mounts inside data disks when
filesystem snapshots are enabled.

16.3 Multiple configurations with filesystem snapshots
------------------------------------------------------

Snapshot support cannot be safely shared by multiple independent
SnapRAID configurations using the same filesystem snapshot namespace.

On Btrfs and Bcachefs, only one SnapRAID configuration with snapshot
support enabled may use data directories within the same subvolume.
On ZFS, the same restriction applies to a dataset, and on NTFS to a volume.

Multiple configurations sharing the same snapshot namespace use the same
`stable`, `pending`, and `scan` snapshots. Running one configuration may
therefore replace or delete recovery snapshots required by another.

16.4 In-place file modifications without timestamp changes
----------------------------------------------------------

SnapRAID detects file modifications by comparing file sizes and
modification timestamps (mtime).

If an application modifies a file in-place without altering its size
or updating its modification timestamp, "diff" and "sync" will not
detect that the file has changed, and its parity will not be updated.

To ensure SnapRAID recognizes such changes, update the file's
modification timestamp (for example, with the "touch" utility).
Otherwise, the difference will only be detected as a hash mismatch
during a subsequent "scrub" or "check".

16.5 Windows non-ASCII case matching
------------------------------------

SnapRAID uses standard, predictable UTF-8 encoding across all platforms,
limiting case-insensitive matching to ASCII characters.

Windows filesystems suffer from the legacy design of enforcing case
insensitivity across Unicode filenames via arbitrary, system-dependent
folding tables. Consequently, Windows obscures the reality of the
filesystem by treating completely different Unicode byte sequences as if
they were the same file.

For example, `À.txt` and `à.txt` are distinct Unicode names, yet Windows
illogically treats them as identical, while SnapRAID preserves their true
distinction.

This limitation is caused by Windows's inconsistent handling of Unicode
filenames and its insistence on case-insensitive aliasing. To prevent
discrepancies arising from Windows's case folding, always specify the
exact case for any non-ASCII characters in configuration paths and filters.

16.6 Partial fix recovery
-------------------------

When `fix` is restricted using `-S, --start` or `-B, --count`,
SnapRAID operates only on the selected blocks and does not perform
file-level finalization or quarantine.

If recovery produces best-effort data that cannot be proven to represent
the current contents, such blocks may still be written under the normal
filename. The uncertainty associated with these recovered blocks is not
persisted after the command terminates.

As a consequence, a later `fix` will treat these readable blocks as valid
file data and will not attempt to recover them again. This can cause a
subsequent recovery to report success while leaving incomplete or stale
data in the file.

The `-S` and `-B` options are intended only for advanced manual recovery,
typically after a normal `fix` has failed. If a partial `fix` reports
unrecoverable data, remove or rename the affected files before attempting
another recovery, or verify their contents manually.

16.7 NTFS hard links and excluded paths
---------------------------------------

On NTFS, directory entries cache file size and timestamps independently
for each hard link. When a file is modified through one link, the directory
entries of other links pointing to the same file may retain stale metadata
until they are accessed.

SnapRAID automatically detects and normalizes this condition whenever at
least two hard links to the same file are encountered during a scan.
However, if only one hard link is scanned while the others are excluded by
filters or reside outside the data disk path, SnapRAID cannot detect the
inode collision and relies on the cached metadata of the single visible
link.

Consequently, modifications made through an excluded or external hard link
may go unnoticed by SnapRAID until the metadata of the included link is
refreshed by the operating system (such as when the file is accessed or
opened through that path).

To avoid this limitation, avoid modifying protected files through hard
links that are excluded from SnapRAID, or ensure that all hard links to a
file are included in the scan so SnapRAID can normalize them.

16.8 Recovery after an interrupted multi-parity sync
----------------------------------------------------

In an array with multiple parity disks, an interrupted `sync` can leave
parity levels temporarily out of sync with each other. For example,
parity 1 might already contain newly updated data, while parity 2 still
contains the old data.

During `check` or `fix`, SnapRAID tries different parity combinations
and validates reconstructed data against the known block hashes (covering
the full block, typically 256 KiB).

If SnapRAID attempts a reconstruction using an inconsistent mix of old
and new parity, the linear RAID equations will produce incorrect data.
For this invalid combination to be falsely accepted, the equations would
have to accidentally reconstruct at least one complete 256 KiB block so
perfectly that its hash matches, while other unsynchronized blocks in the
same stripe are reconstructed with incorrect data.

This scenario does not depend on a hash collision; rather, the mixed
parity equations themselves would have to coincidentally yield the exact
original 256 KiB block contents. While mathematically possible with
specially crafted or highly repetitive data, it is practically impossible
with real-world data.

If all disks are healthy, completing the interrupted `sync` before making
further changes removes this ambiguity. If a disk has failed and recovery
is required, avoid modifying any data disks before running `check` or
`fix`.

16.9 Reserved file and directory names
--------------------------------------

SnapRAID reserves some file, directory, and snapshot names for its internal
operation. Do not use these names for user data:

.snapraidignore - Local exclusion-rules file.
.snapraid - Snapshot container directory.
snapraid-* - SnapRAID-managed ZFS snapshot names.
*.unrecoverable - Damaged files quarantined during recovery.
CONTENT.tmp - Temporary file for content updates.
CONTENT.lock - Content lock file.
PARITY.spaceholder - Temporary space-holder file during Windows parity
    allocation.

Using these names for unrelated data may cause files or snapshots to be
ignored, misinterpreted as metadata, overwritten, renamed, or deleted.

On Windows, all case variants are reserved as well, even in NTFS directories
configured for case-sensitive operation (for example, .SNAPRAIDIGNORE,
.SnapRaid, CONTENT.TMP, and CONTENT.Lock).


17 COPYRIGHT
============

This file is Copyright (C) 2026 Andrea Mazzoleni


18 SEE ALSO
===========

snapraid_log(1), snapraidd(1), rsync(1)
