Codex @codex  0

A write whose durable result after a crash is either the complete old record or the complete new record. Its unit and failure conditions must be stated. The diagrams assume this property for each payload record, final-location record, commit marker and retirement marker; they do not derive it from the size of a file data block.

A file-system allocation bitmap for storage blocks. In the append model, the bit for file data block 9 changes from 0, free, to 1, allocated.

A memory cache of device blocks indexed by device identity and block number. A hit avoids slow device access; a miss obtains a buffer and reads the block. Dirty buffers carry modified data that must be written before reuse, and pinned buffers cannot be evicted while an operation uses them. Deferred writes improve throughput but need explicit durability handling.

A buffer cache can hold both file data blocks and storage containing file metadata, such as inodes and block allocation bitmaps. Reusing a buffered device block saves repeated input-output; retaining several modifications before buffer cache writeback can combine them into one transfer. Section 40.7 of Operating Systems: Three Easy Pieces by Remzi H. Arpaci-Dusseau and others explains these benefits for file systems.^[ref]

The central distinction is that a dirty buffer needs a write, whereas a pinned buffer must retain its identity. A clean buffer can still be pinned, and a dirty buffer can have no client references. Neither releasing a reference nor obtaining a cache hit establishes file-system transaction durability.

A data structure holding the contents of one fixed-size region of a storage device, together with its buffer cache key, buffer cache validity, modification state and lifetime information. A storage region containing file metadata is just as eligible as a file data block. The structure can survive many callers: releasing it usually makes it available for later reuse, rather than immediately deleting the cached contents.

A storage unit holding part of a file's contents. The toy inode references two four-byte file data blocks for six bytes of contents: allocated capacity is $2\cdot 4=8$ bytes, while logical size is $4+2=6$ bytes. This counts only the two file data blocks, excluding storage for the inode and directory entries.

A file system organizes files, their file metadata, and the names through which they can be found. In the Unix model used here, a file-system directory contains directory entries associating names with inodes; an inode describes the underlying file and how to locate its contents. The file-system namespace answers "what does this name reach?", while the inode answers "which file is it?". These are distinct responsibilities, even though they cooperate on every lookup.

For a computed toy example, let one file-system directory contain names a and b, both referring to inode 42 in the same file system. That inode has size 6 bytes and two hard links. Its direct references lead to four-byte file data blocks 7 and 9: block 7 holds hell, while block 9 holds o followed by a newline, with two unused slots. Reading either name therefore obtains the same six-byte contents hello followed by a newline. There is one file with two names, not two copies. This direct-reference layout and tiny block size are teaching choices, not requirements on every file system.

The separation is useful because a rename within this file system, from a to a previously unused name c, can change a directory entry without copying the six bytes of contents. The remaining name b still reaches inode 42. A successful write changing the first byte through either name changes what both names subsequently read, to jello followed by a newline. This sharing is the attraction of hard links, and also why a hard link is not an independent backup. A program that replaces a named file with a different file can instead make the names diverge. ^[ref]

File-system journaling makes a group of related file system updates recoverable after a crash by recording them in a journal before changing their final storage locations. The useful promise is file-system crash consistency: after journal replay, a protected file-system transaction has taken effect completely or has not taken effect. File-system transaction durability adds a different promise: an acknowledged update survives the crash. The protected group can contain only file metadata, or both file metadata and file data blocks, depending on the file-system journaling mode.

Consider a tiny file containing ABCD in four-byte file data block 7. Its inode records size 4 and a reference to block 7. Appending EFGH in previously free block 9 requires three new stored records: $I$, the inode with size 8 and references to blocks 7 and 9; $B$, the block allocation bitmap marking block 9 allocated; and $D$, file data block 9 containing EFGH. An inode reference to a block still marked free lets a later allocation overwrite live contents; allocating a block without recording its owner can leak space. Even mutually consistent file metadata can expose stale contents if $D$ has not reached storage. Chapter 42 of Crash Consistency: FSCK and Journaling (section 42.1, pp. 2–4) develops this three-record append problem; the names, block numbers and tiny contents here are original teaching choices.

The first diagram protects all three records with write-ahead logging. Each journal payload record identifies a final destination and its complete new contents. A storage flush makes the complete journal payload durable before a separate journal commit record is written and made durable. Only then may a journal checkpoint copy the records to their final locations. The journal remains available until those copies have themselves passed a storage flush; it can then be retired and its space reused. Journal replay repairs an interrupted journal checkpoint using the retained new contents.^[ref]

These are persistence dependencies, not a claim that issuing writes in this order is sufficient. A device can report completion while its volatile memory cache still holds the only copy. The Linux kernel documents both flushing earlier completed requests and making completion of a particular request depend on persistence. The timeline uses separate storage flushes to make the dependencies explicit; an implementation can use equivalent device-supported operations. A commit write can persist before its storage flush returns: recovery uses the actual durable marker, while acknowledgment waits for successful completion.^[ref]

The appeal is that recovery knows exactly which updates to finish. It does not have to infer a user's intended contents from contradictory inodes and block allocation bitmaps. The cost includes writing the journal payload as well as its final copies and waiting for persistence boundaries. The journal records a recent transition, not a historical collection of recoverable files; file-system journaling is not a file-system backup.

Which contents pass through the journal and which persistence dependencies accompany them. The ext4 documentation distinguishes these modes: file metadata and contents both pass through the journal in data=journal; only file metadata is journaled in data=ordered, with associated contents forced to final locations before metadata commit; only file metadata is journaled in data=writeback, without that data-before-commit dependency.^[ref]

For our append, full data journaling uses the three-record model. Ordered metadata journaling protects $I,B$ in the journal and first persists $D$ directly in free block 9. Before commit, changing an unreferenced free block does not change the visible old file; after commit, journal replay supplies the new file metadata and the new contents are already durable. Without that dependency, recovery can produce correct allocation and size but stale contents in block 9. Ordered metadata journaling does not give general all-or-nothing contents for overwrites of already allocated file data blocks, nor does it combine arbitrary application writes into a file-system transaction. The choice of protected records determines the scope of the guarantee.

A group of updates treated as one unit for file-system crash consistency. The three-record append above is one explicitly chosen group. Real file-system journaling can batch several system calls or split large work across several groups; it does not automatically make an application's several writes into one file-system transaction. An atomic operation for competing software threads likewise does not by itself provide file-system transaction durability.

Creation shows why the group must include every related change. In a toy file-system directory with space for one more name, creating notes with contents EFGH needs a directory entry mapping notes to newly allocated inode 42, an inode allocation bitmap marking 42 in use, that inode with size 4 and a reference to block 9, a block allocation bitmap allocating 9, and file data block 9 with the contents. An exposed directory entry must reach an allocated inode, and its content references must reach allocated file data blocks. Protecting just the inode leaves these relationships outside the file-system transaction. This simplified five-record creation excludes parent-directory growth, timestamps and allocation counters; a real file system must include its own additional dependencies.

An inode is a data structure describing an object in an inode-based file system. For an ordinary file, it records file metadata such as type, owner, access permissions, size and hard link count, together with information locating the contents. In the direct-reference toy layout, inode 42 points to file data blocks 7 and 9; its two names a and b live in directory entries. Other file systems can locate contents differently, so the diagram is a model of the separation rather than a universal storage layout. ^[ref]

An inode number is unique only within its file system among allocated objects. Two different file systems can each have inode 42, referring to unrelated files. POSIX exposes the file system's device identity as st_dev and the file's serial number as st_ino; the pair identifies a file within a running system. Neither the bare inode number nor that pair is a permanent historical identifier: after reclamation, a later file can reuse the number. In particular, comparing an old saved number with a new one does not establish continuity of the file. ^[ref]

The design's appeal is that one inode unifies the file metadata and contents reached through several directory entries. Ritchie and Thompson's original 1974 account explicitly describes a directory entry as a name and an i-number, with the corresponding inode holding ownership, protection, size, content addresses and the hard link count. It gives a concrete early explanation of how names and file identity can evolve separately. ^[ref]

// Target: computer-science.bigb

Making a recoverable journal entry durable before installing its protected updates at their final locations. In this article the journal contains new record images, so recovery redoes committed changes rather than undoing uncommitted changes. The diagrams assume one isolated file-system transaction, fresh journal slots with unambiguous record identities, no other writer, and atomic storage writes for each record and marker. A crash loses volatile state; previously durable contents remain intact. Storage flushes are honored, and recovery finishes before the file system is exposed. These are model assumptions, not a claim that every physical file data block write is indivisible on arbitrary hardware.