GLOSSARY
What is metadata management?
Metadata management is the practice of keeping an accurate, current description of every file an organization holds, so files can be found and acted on without opening them. It combines what the storage system records with attributes that have to be added. This page covers what a metadata catalog is, how a catalog gets built, how a schema gets designed, and what a catalog does not do.
Home | Resources | Glossary | Metadata and Cataloging
DEFINITION
Metadata management is the practice of describing files so that they can be found, priced, and acted on without opening them. It covers what the storage system already records, such as path, size, owner, and dates, and what has to be added, such as project, retention class, and format detail. For unstructured data, the goal is a single current description of every file, held separately from the files themselves.
KEY TAKEAWAYS
76%
of organizations have made business decisions without consulting available data, because reaching it was too difficult
29
metadata enrichment plugins in Diskover
1
catalog across every storage system you run
What a metadata catalog is.
A catalog describes files without holding them. These four questions separate it from a data catalog and a data fabric, and from the data discovery for unstructured data work it makes possible.
Data catalog compared with metadata catalog:
| DATA CATALOG | METADATA CATALOG | |
|---|---|---|
| What it describes | Structured assets, such as databases, tables, and schemas | Every file an organization holds |
| Where the description comes from | What a database already records about itself | System metadata, plus attributes added by enrichment |
| Scope | Databases | Every storage system, tier, and location |
| Typical attributes | Tables, schemas, and the lineage between them | Attributes such as path, size, owner, and format |
What is a metadata catalog?
A metadata catalog is a searchable index that describes every file an organization holds without storing the files themselves. It records where each file lives, its size, its owner, when it was last opened, and its format, and it keeps that record current as data changes.
How is a metadata catalog different from a data catalog?
A data catalog describes structured assets: databases, tables, schemas, and the lineage between them. A metadata catalog for unstructured data describes files no database is aware of. Most organizations need both, and the two do not overlap.
What is a global file index?
A global file index is one searchable record covering every storage system, tier, and location an organization uses. It is the same idea as a metadata catalog, described from the perspective of coverage rather than content: one place to ask a question, instead of one place per vendor.
What is a data fabric, and how does it relate to a metadata catalog?
A data fabric is an architecture that connects data across many systems through shared metadata and common services, so applications can find and use data without knowing where it sits. For unstructured data, a metadata catalog is one of the parts a data fabric needs: it supplies the current description of every file, such as its path, owner, size, and age. The catalog describes the files, and the rest of the fabric decides how they are reached.
How a metadata catalog gets built.
Building a catalog is an indexing job followed by an enrichment job, and the second one is where the value sits. Enrichment runs as rules at index time rather than as manual typing, and format-aware enrichment plugins read structural detail without reading file contents. For how an indexer, enrichment plugins, and AI share this work, see our post on whether AI can catalog metadata.
What is metadata indexing, and how long does it take?
Metadata indexing is the process of reading file attributes across storage systems and writing them into a searchable index. Duration depends on file count rather than capacity, and on how quickly the storage answers reads. Diskover has indexed 1.4 billion files in 2 hours on three nodes.
What is metadata enrichment?
Metadata enrichment is the addition of attributes a file system does not supply. Format-aware plugins read structural details without reading contents, such as read counts in a sequencing file or codec and duration in a video. Rule-based enrichment adds project, retention class, cost center, and sensitivity signals.
What is the difference between system metadata and enriched metadata?
System metadata is what the file system records automatically: path, name, size, owner, permissions, and dates. Enriched metadata is everything added on top. System metadata tells you a file exists. Enriched metadata tells you what it is for, which is what a decision needs.
Can files carry custom metadata the way objects do?
On some file systems, yes. Dell PowerScale, for example, can store custom key and value pairs on each file, such as project, status, or client, in the same way object storage does. When an indexer reads those values and writes new ones back, the file and the catalog describe the data the same way. Policies can then act on that business context, not only on age or size.
What is data tagging, and why do manual tags fail at scale?
Data tagging attaches labels to files so they can be grouped and acted on. Tags typed by people stop being accurate the moment the estate grows past what a person can review, usually at a few million files. Rules applied at index time keep working, because they are derived rather than entered.
How a metadata schema gets designed.
A catalog is only as useful as the fields everyone agrees on, and the same fields decide what counts as AI-ready data. These four questions cover how the fields get decided and who decides them, which fields to record, how tags differ from fields, and how the values stay consistent.
What is a metadata schema or taxonomy?
A metadata schema is the agreed set of attributes an organization will record, with allowed values for each. A taxonomy is the structure those values sit in. Designing this before enrichment starts is what makes a catalog queryable later, and it is the step most often skipped.
Which fields should a file metadata schema include?
Start with the fields every decision needs, such as project, cost center, retention class, and sensitivity signal, and add a field only when a named decision depends on it. System metadata already supplies attributes such as path, size, owner, and dates, so the schema covers what the storage system does not record. Each field needs a short list of allowed values, because open text entries cannot be queried reliably across a data estate. A schema of a few well-defined fields is more useful than a long list nobody keeps current.
What is the difference between a tag and a metadata field?
A metadata field holds one value from a defined list, such as a retention class, and every file has a place for it. A tag is a label that a file either carries or does not, such as approved or archive candidate, and one file can carry several. Fields suit attributes every file has, and tags suit decisions made about some files. Both can be applied by rule at index time, so neither depends on manual entry.
How do you keep metadata consistent across teams?
Agree on the schema once, publish the allowed values, and apply fields by rule rather than by manual entry. A rule reads what the path and the directory already imply, such as the project or the department, and writes the same value every time. Teams then review the exceptions instead of typing values. Consistency comes from where the values come from, not from asking people to type the same thing.
What a metadata catalog does not do.
A catalog supplies evidence and does not make decisions. Retention classes and deletion rules belong to unstructured data governance, which is a separate body of work.
Does a metadata catalog store copies of files?
No. The catalog stores descriptions, and the files stay exactly where they are. That is what lets one catalog span storage systems which could not otherwise be queried together, and it is why building a catalog does not require a migration.
Can one catalog span several storage vendors?
Yes, and that is the main reason to build one. A storage vendor tool reports accurately on its own platform and is blind to everything else. Read PowerScale, ONTAP, S3, Azure Blob, and the rest into one index, and a single search reaches the whole data estate at once. That is the only way to decide about the estate rather than one system at a time.
What does a metadata catalog not do?
It does not read file contents, so it cannot classify a document by what it says. It does not set retention policy. It supplies the evidence that a policy was applied. And it does not improve data quality, because a catalog of poorly organized data accurately describes poorly organized data.
Does a metadata catalog replace backup?
No. A backup holds copies of files so they can be restored after loss or damage. A metadata catalog holds a description of files, such as path, size, owner, and age, and no copies of the files themselves. The two work together: the catalog shows what exists and where, which helps decide what needs protecting, and the backup holds the protected copies.
Terms in this category.
Metadata management >
Metadata management is the practice of keeping an accurate, current description of every file an organization holds. It combines what the storage system supplies automatically with attributes that have to be derived, and it keeps both current as the data estate changes.
Data tagging >
Data tagging is the attaching of labels to files so they can be grouped and acted on. Tags can be entered by people or applied by a rule at index time. Rules keep working as an estate grows, where tags entered one by one stop being accurate once there are more files than anyone can review.
Metadata catalog >
A metadata catalog is a searchable index that describes every file an organization holds without storing the files themselves. It records where each file sits, its size, its owner, when it last changed, and its format, and it keeps that record current as the estate changes.
Metadata schema >
A metadata schema is the agreed set of attributes an organization will record about its files, with allowed values for each. Deciding it before enrichment begins is what makes a catalog queryable later, and it is the step most often skipped.
Data catalog >
A data catalog is a catalog of structured assets: databases, tables, schemas, and the lineage between them. It describes what a database already knows about itself, which is why it does not cover files and why most organizations end up needing both kinds of catalog.
System metadata >
System metadata is the set of attributes a file system records automatically: path, name, size, owner, group, permissions, and the created, modified, and last accessed dates. It tells you a file exists and where it sits, and nothing about what it is for.
Metadata enrichment >
Metadata enrichment is the addition of attributes a file system does not supply. Format-aware plugins read structural details without reading contents, such as codec and duration in a video. Rule-based enrichment adds project, retention class, cost center, and sensitivity signals from what the path and the directory already imply.
Extended attributes >
Extended attributes are additional key and value pairs a file system can store alongside a file, outside its standard attributes. Support varies by file system, and the values can be lost when a file is copied to a system that does not support them. That is why the catalog stays the durable record, while key fields written back to the file keep context with the data on systems that support it.
Metadata indexing >
Metadata indexing is the reading of file attributes across storage systems into a searchable index. How long it takes depends on file count rather than capacity, and on how quickly each storage system answers reads.
Global file index >
A global file index is one searchable record covering every storage system, tier, and location an organization uses. It is the same idea as a metadata catalog, described from the perspective of coverage: one place to ask a question, instead of one place per vendor.
Data fabric >
A data fabric is an architecture that connects data across systems through shared metadata and common services, so data can be found and used without knowing where it sits. For unstructured data, a metadata catalog supplies the description of every file that a data fabric depends on.
Diskover for metadata management.
Metadata runs through every part of the platform. Here is where to start looking.
One catalog across every vendor.
One index across PowerScale, ONTAP, S3, Azure Blob, and the rest, built while the data stays in place.
29 format-aware plugins.
29 plugins read structural details without reading contents, so each format gives up the attributes it holds.
One index across every site.
Every site, tier, and cloud read into a unified view, so any team can find what another location holds.
Ask the catalog in plain language.
Diskover’s AI connector lets a model query the catalog directly rather than the files.
Related categories.
SOURCES
- Sisense, 2025: 76% have made business decisions without consulting available data.
- Plugin, feature, connector, and endpoint counts are Diskover’s own, current as of the September 2026 release. The full plugin list is published here.
Method note. Platform counts on this page are Diskover’s own and are reported per release, so they move. The 29 figure counts metadata enrichment plugins specifically, which is a subset of the full plugin suite of 55+. The indexing figure is measured on a three-node deployment and scales with file count rather than capacity. The data access figure comes from a 2025 survey of 500 respondents, most of them organizations with 100 to 1,000 employees, so it reports what they say about themselves.
Last updated: October 30, 2026.
See it on your own data.
We’ll index a sample of your environment and show you
what’s recoverable against your current storage spend.