File Systems and Zoned Block Devices
The dm-zoned device mapper target allows using any file system with host managed zoned block devices by hiding the device sequential write constraints. This is a simple solution to enable a file system use but not necessarily the most efficient due to the potentially high overhead of a block based zone reclaim process.
Supporting zoned block devices directly in a file system implementation can lead to a more efficient zone reclaim processing as the file system metadata and file abstraction provide more information on the usage and validity status of storage blocks compared to the raw block device based approach.
Furthermore, a file system design may lend itself well to the sequential write constraint of host managed zoned block devices. This is the case for log-structured file systems such as f2fs and copy-on-write (CoW) file systems such as Btrfs.
zonefs is a very simple file system exposing each zone of a zoned block device as a file. zonefs is included with the upstream Linux kernel since version 5.6.0.
Unlike a regular POSIX-compliant file system with native zoned block device support (e.g. f2fs), zonefs does not hide the sequential write constraint of zoned block devices to the user. Files representing sequential write zones of the device must be written sequentially starting from the end of the file (append only writes).
As such, zonefs is in essence closer to a raw block device access interface than to a full-featured POSIX file system. The goal of zonefs is to simplify the implementation of zoned block device support in applications by replacing raw block device file accesses with the richer regular file API, avoiding relying on direct block device file ioctls which may be more obscure to developers. One example of this approach is the implementation of LSM (log-structured merge) tree structures (such as used in RocksDB and LevelDB) on zoned block devices by allowing SSTables to be stored in a zone file similarly to a regular file system rather than as a range of sectors of the entire disk. The introduction of the higher level construct "one file is one zone" can help reducing the amount of changes needed in the application as well as introducing support for different application programming languages.
The files representing zones are grouped by zone type, which are themselves represented by sub-directories. This file structure is built entirely using zone information provided by the device and so does not require any complex on-disk metadata structure.
zonefs on-disk metadata is composed only of an immutable super block which
persistently stores a magic number and optional feature flags and values. On
mount, zonefs uses the block layer API function
obtain the device zone configuration and populates the mount point with a
static file tree solely based on this information. File sizes come from the
device zone type and write pointer position managed by the device itself.
The super block is always written on disk at sector 0. The first zone of the
device storing the super block is never exposed as a zone file by zonefs. If
the zone containing the super block is a sequential zone, the
tool always "finishes" the zone, that is, it transitions the zone to a full
state to make it read-only, preventing any data write.
Zone Type Sub-Directories
Files representing zones of the same type are grouped together under the same sub-directory automatically created on mount.
For conventional zones, the sub-directory "cnv" is used. This directory is however created if and only if the device has usable conventional zones. If the device only has a single conventional zone at sector 0, the zone will not be exposed as a file as it will be used to store the zonefs super block. For such devices, the "cnv" sub-directory will not be created.
For sequential write zones, the sub-directory "seq" is used.
These two directories are the only directories that exist in zonefs. Users cannot create other directories and cannot rename nor delete the "cnv" and "seq" sub-directories.
The size of the directories indicated by the
st_size field of
obtained with the
fstat() system calls, indicates the number of
files existing under the directory.
Zone files are named using the number of the zone they represent within the set of zones of a particular type. That is, both the "cnv" and "seq" directories contain files named "0", "1", "2", ... The file numbers also represent increasing zone start sector on the device.
All read and write operations to zone files are not allowed beyond the file
maximum size, that is, beyond the zone size. Any access exceeding the zone
size is failed with the
Creating, deleting, renaming or modifying any attribute of files is not allowed.
The number of blocks of a file as reported by
the size of the file zone, or in other words, the maximum file size.
Conventional Zone Files
The size of conventional zone files is fixed to the size of the zone they represent. Conventional zone files cannot be truncated.
These files can be randomly read and written using any type of I/O operation: buffered I/Os, direct I/Os, memory mapped I/Os (mmap), etc. There are no I/O constraint for these files beyond the file size limit mentioned above.
Sequential zone files
The size of sequential zone files grouped in the "seq" sub-directory represents the file's zone write pointer position relative to the zone start sector.
Sequential zone files can only be written sequentially, starting from the file end, that is, write operations can only be append writes. Zonefs makes no attempt at accepting random writes and will fail any write request that has a start offset not corresponding to the end of the file, or to the end of the last write issued and still in-flight (for asynchronous I/O operations).
Since dirty page writeback by the page cache does not guarantee a sequential write pattern, zonefs prevents buffered writes and writeable shared mappings on sequential files. Only direct I/O writes are accepted for these files. zonefs relies on the sequential delivery of write I/O requests to the device implemented by the block layer elevator (See Write Command Ordering).
There are no restrictions on the type of I/O used for read operations in sequential zone files. Buffered I/Os, direct I/Os and shared read mappings are all accepted.
Truncating sequential zone files is allowed only down to 0, in which case, the zone is reset to rewind the file zone write pointer position to the start of the zone, or up to the zone size, in which case the file's zone is transitioned to the FULL state (finish zone operation).
Several optional features of zonefs can be enabled at format time.
- Conventional zone aggregation: ranges of contiguous conventional zones can be aggregated into a single larger file instead of the default one file per zone.
- File ownership: The owner UID and GID of zone files is by default 0 (root) but can be changed to any valid UID/GID.
- File access permissions: the default 640 access permissions can be changed.
IO error handling
Zoned block devices may fail I/O requests for reasons similar to regular block devices, e.g. due to bad sectors. However, in addition to such known I/O failure pattern, the standards governing zoned block devices behavior define additional conditions that can result in I/O errors.
A zone may transition to the read-only condition: While the data already written in the zone is still readable, the zone can no longer be written. No user action on the zone (zone management command or read/write access) can change the zone condition back to a normal read/write state. While the reasons for the device to transition a zone to read-only state are not defined by the standards, a typical cause for such transition would be a defective write head on an HDD (all zones under this head are changed to read-only).
A zone may transition to the offline condition: An offline zone cannot be read nor written. No user action can transition an offline zone back to an operational good state. Similarly to zone read-only transitions, the reasons for a drive to transition a zone to the offline condition are undefined. A typical cause would be a defective read-write head on an HDD causing all zones on the platter under the broken head to be inaccessible.
Unaligned write errors: These errors result from the host issuing write requests with a start sector that does not correspond to a zone write pointer position when the write request is executed by the device. Even though zonefs enforces sequential file write for sequential zones, unaligned write errors may still happen in the case of a partial failure of a very large direct I/O operation split into multiple BIOs/requests or asynchronous I/O operations. If one of the write request within the set of sequential write requests issued to the device fails, all write requests queued after it will become unaligned and fail.
Delayed write errors: Similarly to regular block devices, if the device side write cache is enabled, write errors may occur in ranges of previously completed writes when the device write cache is flushed, e.g. on
fsync(). Similarly to the previous immediate unaligned write error case, delayed write errors can propagate through a stream of cached sequential data for a zone causing all data to be dropped after the sector that caused the error.
All I/O errors detected by zonefs are notified to the user with an error code return for the system call that triggered or detected the error. The recovery actions taken by zonefs in response to I/O errors depend on the I/O type (read vs write) and on the reason for the error (bad sector, unaligned writes or zone condition change).
For read I/O errors, zonefs does not execute any particular recovery action, but only if the file zone is still in a good condition and there is no inconsistency between the file inode size and its zone write pointer position. If a problem is detected, I/O error recovery is executed (see below table).
For write I/O errors, zonefs I/O error recovery is always executed.
A zone condition change to read-only or offline also always triggers zonefs I/O error recovery.
zonefs minimal I/O error recovery may change a file size and file access permissions.
File size changes: Immediate or delayed write errors in a sequential zone file may cause the file inode size to be inconsistent with the amount of data successfully written in the file zone. For instance, the partial failure of a multi-BIO large write operation will cause the zone write pointer to advance partially, even though the entire write operation will be reported as failed to the user. In such case, the file inode size must be advanced to reflect the zone write pointer change and eventually allow the user to restart writing at the end of the file. A file size may also be reduced to reflect a delayed write error detected on fsync(): in this case, the amount of data effectively written in the zone may be less than originally indicated by the file inode size. After such I/O error, zonefs always fixes the file inode size to reflect the amount of data persistently stored in the file zone.
Access permission changes: A zone condition change to read-only is indicated with a change in the file access permissions to render the file read-only. This disables changes to the file attributes and data modification. For offline zones, all permissions (read and write) to the file are disabled.
Further action taken by zonefs I/O error recovery can be controlled by the user with the "errors=xxx" mount option. The table below summarizes the result of zonefs I/O error processing depending on the mount option and on the zone conditions.
|"errors=xxx"mount option||device zonecondition||filesize||fileread||filewrite||deviceread||devicewrite|
- The "errors=remount-ro" mount option is the default behavior of zonefs I/O error processing if no errors mount option is specified.
- With the "errors=remount-ro" mount option, the change of the file access permissions to read-only applies to all files. The file system is remounted read-only.
- Access permission and file size changes due to the device transitioning zones to the offline condition are permanent. Remounting or reformatting the device with mkfs.zonefs (mkzonefs) will not change back offline zone files to a good state.
- File access permission changes to read-only due to the device transitioning zones to the read-only condition are permanent. Remounting or reformatting the device will not re-enable file write access.
- File access permission changes implied by the remount-ro, zone-ro and zone-offline mount options are temporary for zones in a good condition. Unmounting and remounting the file system will restore the previous default (format time values) access rights to the files affected.
- The repair mount option triggers only the minimal set of I/O error recovery actions, that is, file size fixes for zones in a good condition. Zones indicated as being read-only or offline by the device still imply changes to the zone file access permissions as noted in the table above.
zonefs define the "errors=
- remount-ro (default)
The run-time I/O error actions defined for each behavior are detailed in the previous section. Mount time I/O errors will cause the mount operation to fail. The handling of read-only zones also differs between mount-time and run-time. If a read-only zone is found at mount time, the zone is always treated in the same manner as offline zones, that is, all accesses are disabled and the zone file size set to 0. This is necessary as the write pointer of read-only zones is defined as invalib by the ZBC and ZAC standards, making it impossible to discover the amount of data that has been written to the zone. In the case of a read-only zone discovered at run-time, as indicated in the previous section. the size of the zone file is left unchanged from its last updated value.
Zonefs User Space Tools
mkzonefs tool is used to format zoned block devices for use with zonefs.
This tool is available on
zonefs-tools also includes a test suite which can be run against any zoned block device, including nullblk block device created with zoned mode.
The following formats a 15TB host-managed SMR HDD with 256 MB zones with the conventional zones aggregation feature enabled::
# mkzonefs -o aggr_cnv /dev/sdX # mount -t zonefs /dev/sdX /mnt # ls -l /mnt/ total 0 dr-xr-xr-x 2 root root 1 Nov 25 13:23 cnv dr-xr-xr-x 2 root root 55356 Nov 25 13:23 seq
The size of the zone files sub-directories indicate the number of files existing for each type of zones. In this example, there is only one conventional zone file (all conventional zones are aggregated under a single file).
# ls -l /mnt/cnv total 137101312 -rw-r----- 1 root root 140391743488 Nov 25 13:23 0
This aggregated conventional zone file can be used as a regular file::
# mkfs.ext4 /mnt/cnv/0 # mount -o loop /mnt/cnv/0 /data
The "seq" sub-directory grouping files for sequential write zones has in this example 55356 zones::
# ls -lv /mnt/seq total 14511243264 -rw-r----- 1 root root 0 Nov 25 13:23 0 -rw-r----- 1 root root 0 Nov 25 13:23 1 -rw-r----- 1 root root 0 Nov 25 13:23 2 ... -rw-r----- 1 root root 0 Nov 25 13:23 55354 -rw-r----- 1 root root 0 Nov 25 13:23 55355
For sequential write zone files, the file size changes as data is appended at the end of the file, similarly to any regular file system::
# dd if=/dev/zero of=/mnt/seq/0 bs=4096 count=1 conv=notrunc oflag=direct 1+0 records in 1+0 records out 4096 bytes (4.1 kB, 4.0 KiB) copied, 0.00044121 s, 9.3 MB/s # ls -l /mnt/seq/0 -rw-r----- 1 root root 4096 Nov 25 13:23 /mnt/seq/0
The written file can be truncated to the zone size, preventing any further write operation::
# truncate -s 268435456 /mnt/seq/0 # ls -l /mnt/seq/0 -rw-r----- 1 root root 268435456 Nov 25 13:49 /mnt/seq/0
Truncation to 0 size allows freeing the file zone storage space and restart append-writes to the file::
# truncate -s 0 /mnt/seq/0 # ls -l /mnt/seq/0 -rw-r----- 1 root root 0 Nov 25 13:49 /mnt/seq/0
Since files are statically mapped to zones on the disk, the number of blocks of a file as reported by stat() and fstat() indicates the size of the file zone::
# stat /mnt/seq/0 File: /mnt/seq/0 Size: 0 Blocks: 524288 IO Block: 4096 regular empty file Device: 870h/2160d Inode: 50431 Links: 1 Access: (0640/-rw-r-----) Uid: ( 0/ root) Gid: ( 0/ root) Access: 2019-11-25 13:23:57.048971997 +0900 Modify: 2019-11-25 13:52:25.553805765 +0900 Change: 2019-11-25 13:52:25.553805765 +0900 Birth: -
The number of blocks of the file ("Blocks") in units of 512B blocks gives the maximum file size of 524288 * 512 B = 256 MB, corresponding to the device zone size in this example. Of note is that the "IO block" field always indicates the minimum I/O size for writes and corresponds to the device physical sector size.
The Flash-Friendly File System (f2fs) was designed on a basis of a log-structured file system approach but modified to avoid the classical problems of the traditional log-structured approach (e.g. The snowball effect of wandering trees and the high cleaning overhead).
f2fs supports various parameters not only for configuring on-disk layout but also for selecting allocation and cleaning algorithms.
Zoned Block Device Support
Zoned block device support was added to f2fs with kernel 4.10. Since f2fs uses a metadata block on-disk format with fixed block location, only zoned block devices which include conventional zones can be supported. Zoned devices composed entirely of sequential zones cannot be used with f2fs as a standalone device and require a multi-device setup to place metadata blocks on a randomly writable storage. f2fs supports multi-device setup where multiple block device address spaces are linearly concatenated to form a logically larger block device. The dm-linear device mapper target can also be used to create a logical device composed of conventional zones and sequential zones suitable for f2fs.
f2fs zoned block device support was achieved using the following principles.
- Section Alignment In f2fs, a section is a group of fixed size segments (2 MB). The number of segments in a section is determined to match the zoned device zone size. For instance, with a 256 MB zone size, a section contains 128 segments of 2MB.
- Forced LFS mode By default, f2fs tries to optimize block allocation to avoid excessive append write by allowing some random writes within segments. The LFS mode forces sequential writes to segments and the sequential use of segments within sections, resulting in full compliance with zoned block devices write constraint.
- Zone reset as discard operation Block discard (or trim) used to indicate to a device that a block or range of blocks are no longer in use is replaced with execution of a zone write pointer reset command when all blocks of all segments of a section are free, allowing the section to be reused.
Compared to a solution using the dm-zoned device mapper target, performance of f2fs on zoned devices does not suffer from zone reclaim overhead as writes are always sequential and do not require on-disk temporary buffering. f2fs garbage collection (segment cleanup) will generate overhead only for workloads frequently deleting file or modifying files data.
Zone Capacity Support
NVMe ZNS SSDs can have a per zone capacity that is smaller than the zone size. To support ZNS devices, f2fs ensures that block allocation and accounting only considers the blocks in a zone that are within the zone capacity. This support for NVMe ZNS zone capacity is available since Linux kernel version 5.10.
Additionally, f2fs volumes need some storage space that is randomly writable to store and update in-place metadata blocks for the volume. Since NVMe zoned namespaces do not have conventional zones, a f2fs volume cannot be self-contained within a single NVMe zoned namespace. To format a f2fs volume using a NVMe zoned namespace, a multi-device volume format must be used to provide an additional regular block device to store the volume metadata blocks. This additional regular block device can be either a regular namespace on the same NVMe device or a regular namespace on another NVMe device.
f2fs uses 32-bits block numbers with a block size of 4 KB. This results in a maximum volume size of 16 TB. Any device or combination of devices (for a multi-device volume) with a total capacity larger than 16 TB cannot be used with f2fs.
To overcome this limit, the dm-linear device mapper target can be used to partition a zoned block device into serviceable smaller logical devices. This configuration must ensure that each logical device created is assigned a sufficient amount of conventional zones to store f2fs fixed location metadata blocks.
Usage Example with a Host Managed SMR HDD
To format a zoned block device with mkfs.f2fs, the option
-m must be
# mkfs.f2fs -m /dev/sdb f2fs-tools: mkfs.f2fs Ver: 1.12.0 (2018-11-12) Info: Disable heap-based policy Info: Debug level = 0 Info: Trim is enabled Info: [/dev/sdb] Disk Model: HGST HSH721415AL Info: Host-managed zoned block device: 55880 zones, 524 randomly writeable zones 65536 blocks per zone Info: Segments per section = 128 Info: Sections per zone = 1 Info: sector size = 4096 Info: total sectors = 3662151680 (14305280 MB) Info: zone aligned segment0 blkaddr: 65536 Info: format version with "Linux version 5.0.16-300.fc30.x86_64 (email@example.com) (gcc version 9.1.1 20190503 (Red Hat 9.1.1-1) (GCC)) #1 SMP Tue May 14 19:33:09 UTC 2019" Info: [/dev/sdb] Discarding device Info: Discarded 14305280 MB Info: Overprovision ratio = 0.600% Info: Overprovision segments = 86254 (GC reserved = 43690) Info: format successful
The formatted zoned block device can now be directly mounted without any other setup necessary.
# mount /dev/sdb /mnt
Usage Example with a NVMe ZNS SSD
Unlike SMR hard-disks, the kernel does not select by default the mq-deadline block IO scheduler for block devices representing NVMe zoned namespaces. To ensure that the regular write operations used by f2fs are delivered to the device in sequential order, the IO scheduler for the NVMe zoned namespace block device must be set to mq-deadline. This is done with the following command.
# echo mq-deadline > /sys/block/nvme1n1/queue/scheduler
Where /dev/nvme1n1 is the block device file of the zoned namespace that will be
used for the f2fs volume. Using this namespace, a multi-device f2fs volume
using an additional regular block device (
/dev/nvme0n1 in the following
example) can be formatted using the -c option of mkfs.f2fs, as shown in the
# mkfs.f2fs -f -m -c /dev/nvme1n1 /dev/nvme0n1 F2FS-tools: mkfs.f2fs Ver: 1.14.0 (2021-06-23) Info: Disable heap-based policy Info: Debug level = 0 Info: Trim is enabled Info: Host-managed zoned block device: 2048 zones, 0 randomly writeable zones 524288 blocks per zone Info: Segments per section = 1024 Info: Sections per zone = 1 Info: sector size = 4096 Info: total sectors = 1107296256 (4325376 MB) Info: zone aligned segment0 blkaddr: 524288 Info: format version with "Linux version 5.13.0-rc6+ (user1@brahmaputra) (gcc (Ubuntu 10.3.0-1ubuntu1) 10.3.0, GNU ld (GNU Binutils for Ubuntu) 2.36.1) #2 SMP Fri Jun 18 16:45:29 IST 2021" Info: [/dev/nvme0n1] Discarding device Info: This device doesn't support BLKSECDISCARD Info: This device doesn't support BLKDISCARD Info: [/dev/nvme1n1] Discarding device Info: Discarded 4194304 MB Info: Overprovision ratio = 3.090% Info: Overprovision segments = 74918 (GC reserved = 40216) Info: format successful
To mount the volume formatted with the above command, the regular block device must be specified.
# mount -t f2fs /dev/nvme0n1 /mnt/f2fs/
Btrfs is a file system based on the copy-on-write (CoW) principle resulting in any block update to never be written in-place. Work is ongoing to add native ZBD support by changing the block allocation algorithm and block IO issuing code.
Block Allocation Changes
Btrfs block management relies on grouping of blocks into block groups, with each group composed of one or more device extent. The device extents of a block group may belong to different devices (e.g. In the case of a RAID volume). ZBD support changes the default device extent size to the size of the device zones so that all device extents are always aligned to a zone.
Allocation of blocks within a block group is changed so that the allocation is always sequential from the beginning of the block group. To do so, an allocation pointer is added to block groups and used as the allocation hint. The changes also ensure that block freed below the allocation pointer are ignored, resulting in sequential block allocation within each group regardless of the block group usage.
While the introduction of the allocation pointer ensures that blocks are allocated sequentially within groups, so sequentially within zones, I/Os to write out newly allocated blocks may be issued out of order causing errors when writing to sequential zones. This problem is solved by introducing a write I/O request staging list to each block group. This list is used to delay the execution of unaligned write requests within a block group.
The zones of a block group are reset to allow rewriting only when the block group is being freed, that is, when all the blocks within the block group are unused.
For Btrfs volumes composed of multiple disks, restrictions are added to ensure that all disks have the same zone model and in the case of zoned block devices, the same zone size. This matches the existing Btrfs constraint that all device extents in a block group must have the same size.
Btrfs zoned block device support is still in development and will be available in stable releases after the usual upstream review process completes.
XFS currently does not support zoned block devices. The dm-zoned device mapper target must be used to enable zoned device use with XFS.
An early design document discussed the development work necessary to support host aware and host managed disks with XFS. Parts of this design have already been implemented and included into the kernel stable releases (e.g. Per inode reverse block mapping b-trees feature). However, more work is necessary to fully support zoned block devices.
This article describes attempts at improving ext4 performance with host aware zoned block devices using changes to the file system journal management. The changes are small and succeed in maintaining good performance. However, support for host managed zoned block devices is not provided as some fundamental ext4 design aspects cannot be easily changed to match host managed device constraints.
These optimizations for host aware zoned block devices is a research work and is not included in ext4 stable kernel releases. ext4 also does not support host managed disks. Similarly to XFS, the ext4 file system can however be used together with the dm-zoned device mapper target.