Explore our new Cybersecurity Terms Glossary, explained by our experts.

View the Glossary View the Glossary
PRIVIA

Cybersecurity Terms Glossary

What Is File System Forensics?

File system forensics reads the structures that organise a disk — NTFS metadata, ext4 inodes, APFS snapshots — to establish which files existed, when they changed, and what has since been deleted.

Illustration for the cybersecurity term File System Forensics
Home / Blog / Cybersecurity Terms / What Is File System Forensics?
17 November 2026

What Is File System Forensics?

What Is File System Forensics?

A file system is a database that happens to store files. It tracks where each file’s content sits on the device, who owns it, when it was created and modified, which blocks are free, and — crucially for investigators — it keeps records of operations in journals designed for crash recovery rather than for auditing. File system forensics is the practice of reading those structures directly instead of trusting what the operating system chooses to display.

NTFS: The One You Will Meet Most

Every object on an NTFS volume has a record in the master file table, including the table itself. Each record holds attributes rather than fixed fields: $STANDARD_INFORMATION with the familiar timestamps and flags, $FILE_NAME with a second set of timestamps and the parent directory reference, and $DATA with the content, which for small files is stored inside the record itself. Deleting a file clears an in-use flag and leaves the record otherwise intact, which is why deleted entries are recoverable long after the fact — explored further in master file table analysis.

Around it sit three artefacts worth knowing by name. The change journal, $UsnJrnl, logs every create, write, rename, and delete with a reason code, giving a compact history of file activity even where the files are gone. The transaction log, $LogFile, holds lower-level records that can reconstruct operations the journal has already aged out. And the directory index, $I30, retains entries for deleted files in its unallocated space, which frequently proves that a file once existed in a specific folder.

Alternate data streams deserve a mention: NTFS allows additional named streams on any file, invisible in a normal listing, and they remain a standard hiding place for payloads and a routine source of the Zone.Identifier marker that shows a file was downloaded from the internet.

ext4 and APFS

On Linux, ext4 stores metadata in inodes, with extents mapping content to disk and a journal recording metadata changes. Deletion zeroes the extent pointers in most configurations, which makes recovery harder than on NTFS and pushes investigators towards file carving and journal analysis. One underused detail is the creation time: it exists in ext4 inodes but is not exposed by stat, and must be read with debugfs.

APFS on macOS is copy-on-write, so modifying a file writes new blocks rather than overwriting old ones, and snapshots are cheap and common. That is an investigative gift: a volume may contain multiple historical states of the same file system, including from before an attacker’s clean-up. Time Machine local snapshots and the FSEvents database add further history at low cost.

Timestamps Need Care

Each file system records different times with different precision, and they are not interchangeable. NTFS keeps four timestamps in two separate attributes, so eight values per file, and the duplication is what exposes timestamp manipulation. Many file systems store times in UTC while others use local time, archives embed their own, and copying a file across file systems rewrites some values and preserves others. Before any timeline work, establish the system’s time zone, check for clock changes in the event log, and state the basis in the report.

Tooling and Output

The Sleuth Kit remains the foundation, with mmls for partition layout, fls for file listings including deleted entries, istat for metadata, and icat for content extraction. Autopsy wraps it in an interface. Eric Zimmerman’s MFTECmd parses NTFS metadata at speed, and plaso builds the super timeline that most investigations actually run on, merging file system events with logs and application artefacts into one chronology.

Work from a verified image rather than the live volume, keep the chain of custody intact, and record tool versions, since parser behaviour on edge cases changes between releases. At Privia Security, we use file system analysis to establish what an intruder staged and when, which is usually the backbone of the timeline in our digital forensic incident response services.

FAQ – Frequently Asked Questions

Can You Prove a File Was Deleted and When?

Often, within limits. The NTFS change journal records delete operations with reason codes and sequence numbers, and directory index slack space can retain entries for removed files. The master file table record itself usually survives with its in-use flag cleared. What you rarely get is a reliable deletion timestamp, so phrase findings as the window within which deletion must have occurred.

Which File System Is Hardest to Investigate?

ext4 is generally the least forgiving, because deletion typically zeroes the extent pointers that map a file to its content, leaving carving as the main recovery route. APFS is the most generous thanks to copy-on-write and frequent snapshots. NTFS sits between the two and has by far the richest set of documented metadata artefacts.

Why Does NTFS Keep Two Sets of Timestamps?

One set lives in the standard information attribute and the other in the file name attribute, which exists for directory indexing. Ordinary applications update the first, while the second is maintained by the file system itself. Tools that falsify timestamps usually change only the first set, so a discrepancy between the two is strong evidence of manipulation.

Do I Need a Full Disk Image for File System Analysis?

For unallocated space, slack, and carving, yes. For metadata analysis you can work from a triage collection containing the master file table, the change journal, the transaction log, and the directory indexes, which is a few hundred megabytes rather than terabytes. That approach scales across dozens of hosts during an intrusion investigation.

Related Terms

You May Be Interested In These

What Is API Security?

Cybersecurity Terms

17 November 2026

What Is API Security?

Read More
What Is Adversarial Machine Learning?

Cybersecurity Terms

17 November 2026

What Is Adversarial Machine Learning?

Read More
What Is an Infostealer?

Cybersecurity Terms

17 November 2026

What Is an Infostealer?

Read More