Account Management and Data Governance · Topic 9 of 9
Trust Center, Horizon Catalog, Universal Search and Lineage
Enable Trust Center scanner packages and roles, and use Horizon Catalog, Universal Search and data lineage to find, govern and trace your data.
- 5 min
- Trust Center roles are the application roles
TRUST_CENTER_VIEWERandTRUST_CENTER_ADMIN, granted byACCOUNTADMIN. - Packages: Security Essentials (always on), CIS Benchmarks, Threat Intelligence, AI Security, Data Security.
- Violations are persistent and remediable; detections are one-time events.
What it is
The Trust Center checks your account against security recommendations. It runs scanners, grouped in scanner packages, and reports findings so an administrator can fix weak configuration such as password-only users without MFA, missing network policies or risky sign-ins.
Snowflake Horizon Catalog is Snowflake's catalog and governance layer for data inside and outside Snowflake. It brings together governance features you have already met (classification, tagging, masking and row access policies, access history, the Trust Center) with discovery and lineage. Universal Search finds objects, and data lineage shows where data came from and what depends on it.
How it works
Trust Center roles
An ACCOUNTADMIN grants one of two application roles: SNOWFLAKE.TRUST_CENTER_VIEWER or SNOWFLAKE.TRUST_CENTER_ADMIN. Viewers can see the Overview, Violations and other findings tabs. Admins can also manage violations and scanners. In the organization account, GLOBALORGADMIN is the role used instead. The Organization tab additionally needs ORGANIZATION_SECURITY_VIEWER. Reader accounts are not supported.
GRANT APPLICATION ROLE SNOWFLAKE.TRUST_CENTER_VIEWER TO ROLE sec_auditor;
GRANT APPLICATION ROLE SNOWFLAKE.TRUST_CENTER_ADMIN TO ROLE sec_ops;
Scanner packages
| Package | Default | What it checks |
|---|---|---|
| Security Essentials | On, cannot be turned off | MFA enrollment for password-using human users, network policies, event table setup |
| CIS Benchmarks | Off | The account against the CIS Snowflake Benchmarks; runs daily by default |
| Threat Intelligence | Off | Password-only sign ins, authentication failures, dormant users, sensitive parameter changes, unusual login IPs |
| AI Security | Off | Cortex related checks such as sensitive data accessed by agents |
| Data Security | Off | Event-driven detections for unusual unloads to external and internal stages |
You enable a package in the Trust Center and it runs immediately. After that, admins can turn individual scanners on or off and change schedules of the schedule-based ones. Event-driven scanners report within about an hour of the event, and some can take up to four hours.
Findings
| Finding | Meaning | What you do |
|---|---|---|
| Violation | A persistent misconfiguration, with remediation guidance | Fix it; it stays listed until the next scanner run |
| Detection | A one-time event such as a suspicious sign in or data transfer | Investigate; it cannot be remediated or managed through a lifecycle |
Violations have a severity of low, medium, high or critical. Email and programmatic notifications can be set up, and resolved violations do not send emails. The Overview tab shows how ready the account is for secure authentication.
A scenario that says "find users who sign in with only a password" points to the Security Essentials or Threat Intelligence package. "Check the account against an industry benchmark" points to CIS Benchmarks.
Horizon Catalog
Horizon Catalog is described as a catalog for data in and outside Snowflake. It covers:
- Connecting data: Iceberg tables across engines, the Iceberg REST Catalog API, external tables, catalog-linked databases and an internal marketplace.
- Context for AI: semantic views, column-level lineage, auto-generated descriptions through Snowflake Cortex and object tagging.
- Governed AI: policies enforced at the query engine, so the same rules apply to analysts, BI tools and AI agents, plus classification, data quality monitoring, access history and the Trust Center.
- Business continuity: replication of databases and accounts.
The point for an administrator: governance is enforced where the data is queried, not in each tool.
Universal Search
Universal Search in Snowsight searches metadata, not data contents: object names, comments and tags on objects and columns. It also finds Marketplace and Internal Marketplace products and documentation. Results depend on the privileges of your active role and secondary roles, so users never see objects they cannot access. New or recreated objects can take a few hours to appear.
If an expected object is missing from search, first check the role in use. Missing access and indexing delay are the usual causes.
Data lineage
The Lineage tab in Snowsight shows upstream sources and downstream consumers of an object, down to the column level. It works for tables, views, materialized views, dynamic tables and similar objects, and records operations such as COPY INTO, CREATE TABLE AS SELECT, INSERT ... SELECT, MERGE and CREATE VIEW. It requires Enterprise Edition or higher.
To view lineage a role needs the account privilege VIEW LINEAGE, which PUBLIC holds by default, plus USAGE on the database and schema and SELECT or REFERENCES on the object. The function SNOWFLAKE.CORE.GET_LINEAGE returns lineage in SQL.
SELECT * FROM TABLE(SNOWFLAKE.CORE.GET_LINEAGE(
'SALES.PUBLIC.ORDERS_SUMMARY', 'TABLE', 'UPSTREAM', 3));
Use lineage for impact analysis ("what breaks if this column changes") and to trace sensitive data through copies.
Common mistakes
Trying to turn off Security Essentials. It is on by default and cannot be disabled.
Treating detections like violations. Detections are one-time events to investigate; there is no remediation or lifecycle.
Expecting the Trust Center to fix issues. It reports and gives guidance; the administrator makes the change.
Expecting Universal Search to search table contents. It searches metadata only.
Expecting lineage on Standard Edition. It needs Enterprise Edition or higher.
What to remember for the exam
- Trust Center roles are the application roles
TRUST_CENTER_VIEWERandTRUST_CENTER_ADMIN, granted byACCOUNTADMIN. - Packages: Security Essentials (always on), CIS Benchmarks, Threat Intelligence, AI Security, Data Security.
- Violations are persistent and remediable; detections are one-time events.
- Universal Search covers metadata and respects the active roles.
- Lineage: column level, Enterprise or higher,
VIEW LINEAGE,SNOWFLAKE.CORE.GET_LINEAGE.
Check yourself
Which Trust Center package is always enabled?
Security Essentials, which checks MFA enrollment, network policies and event table setup.
What is the difference between a violation and a detection?
A violation is a persistent misconfiguration you remediate; a detection is a one-time event you investigate.
What does Universal Search return for a user with limited privileges?
Only objects the active role and secondary roles can access.
Topic 9 of 9 in Account Management and Data Governance
Test yourself on Account Management and Data Governance
The full Snowflake ADA-C02 question bank has 814 questions, with timed practice exams and an explanation for every answer.
This study guide is independent and is not affiliated with, endorsed by, or sponsored by Snowflake. Vendor names, certification names and exam codes are trademarks of their respective owners and are used for identification only.