> ## Documentation Index > Fetch the complete documentation index at: https://docs.nx1cloud.com/llms.txt > Use this file to discover all available pages before exploring further. # DataHub in NexusOne > Overview of DataHub, a central metadata platform for discovery, lineage, and governance in NexusOne. DataHub is a central metadata platform used for discovering datasets, understanding lineage, documenting tables, and managing ownership across an organization. It provides unified visibility into the following: * Dataset metadata * Schema structure * Data lineage * Dashboard and pipeline dependencies * Business glossary and governance information DataHub integrates with multiple systems, including data lakes, data warehouses, BI tools, and orchestration frameworks. ## Accessing DataHub DataHub provides a web-based interface that allows you to browse datasets, view lineage, and manage metadata. You can access it through a standard web browser. ### Accessing the UI DataHub is available at the following designated URL: ```text theme={null} https://datacatalog..nx1cloud.com/ ``` When you purchase NexusOne, you receive a client name. Replace `client` with your assigned client name. ### Authentication and authorization DataHub supports multiple authentication mechanisms, such as: * Single Sign-On (SSO) * OAuth 2.0 or OpenID Connect After navigating to the previously mentioned URL, you must enter the credentials assigned to you when you purchased NexusOne to be successfully authenticated. For authorization, DataHub controls your access to perform specific actions using role-based permissions in KeyCloak. The typical roles include: * **Viewer**: Read-only access to datasets, lineage, and documentation * **Editor**: Read and write access to update descriptions, add tags, and modify owners * **Admin**: Authority over all actions, such as ingestion pipelines, glossary terms, and policies ## Browsing and discovering assets Assets represent items such as datasets, dashboards, pipelines, or domains. DataHub indexes these assets and records who owns them, how they connect to other assets, applied tags, and documentation. ### Searching for assets You can search across all metadata types from the main search bar, such as: * Exact names or partial names, for example `sales_daily` or `sales` * Column names, for example, `customer_id` * Tags or glossary terms * The name of a task or pipeline that produces or transforms a dataset Searching for an asset While searching, you can also filter by the following categories: * **Platforms**: Spark, Airflow, Hive/Iceberg, Trino, or Superset * **Domains**: Sales, Marketing, Finance, or Operations * **Tags**: PII, Sensitive, or Analytics ### Dataset overview page Opening a dataset displays an overview page with several tabs containing details of the metadata. Some of these include: * `Columns`: Column names, data types, descriptions * `Description`: Purpose and usage notes * `Owners`: Responsible team or users * `Tags`: Labels that classify or categorize the dataset * `Lineage`: Upstream and downstream graph * `Properties`: Storage location and system-specific details * `Data Preview`: If enabled, it displays sample rows for quick inspection A dataset overview page ## Metadata details The Metadata details page provides a comprehensive view of a dataset's technical, business, and operational metadata. It's the central place where you can understand what a dataset contains, how it's used, where it came from, and who is responsible for it. ### Dataset metadata summary The summary sidebar provides a high-level summary of a dataset, so you can quickly determine its purpose and context. A dataset metadata summary Some of the details the metadata contains include: * `Documentation`: Human-readable explanation of what the dataset contains * `Owners`: Person or team responsible for maintaining the dataset * `Domain`: Organizational unit or functional area the dataset belongs to, such as Finance, Marketing, or Engineering * `Data Product`: Data managed by a domain * `Tags`: Informational labels such as PII, Finance, Sensitive, or Analytics * `Composed of`: Platform-specific representation of the dataset, such as Iceberg, or Trino * `Status`: Timestamp indicating when someone last modified the metadata ### Schema detail The table that appears when you open a dataset describes a schema detail, as [previously described](#dataset-overview-page). However, the focus is on the schema in the "Columns" tab. Understanding the schema helps you know what data exists, how to use it, and how to interpret it correctly. A dataset metadata summary Some of these details include the following: * Column name * Data types such as `STRING`, `INTEGER`, or `TIMESTAMP` * Business-friendly description containing meanings or definitions * Tags containing labels for classification ### Lineage details The Lineage tab reveals what produced a dataset and which systems consume it. With this, you can identify dependencies, evaluate the impact of changes, and trace issues back to their origin. Two types of lineage exist on DataHub: dataset and job lineage. #### Dataset lineage Dataset-level lineage displays a schema and its table. A dataset lineage To view it, take the following steps: 1. Search for the dataset name, schema, or other identifiers indexed in DataHub. 2. Open the dataset overview page and select the **Lineage** tab. #### Job lineage Job lineage displays the dataset lineage, along with the task or pipeline that produced or transformed it. Some of the supported tasks or pipelines include: * Spark jobs, such as batch or streaming * Airflow DAGs and task-level lineage * SQL-based ETL jobs such as Trino or dbt A job lineage To view it, take the following steps: 1. Search for the job or pipeline name. 2. Open the job overview page and select the **Lineage** tab. ## API usage DataHub provides REST and GraphQL APIs for programmatic metadata updates. These APIs are typically used for automated metadata pipelines, CI/CD workflows, or custom integrations where you need to update metadata without using the UI. ### REST API The REST API supports creating, updating, and querying metadata by sending Metadata Change Events (MCE) or Metadata Aspects. An MCE is a message that describes changes to one or more assets. A Metadata Aspect is a specific piece of metadata about an asset, such as its ownership, tags, or schema. For example, update a dataset description. ```bash theme={null} curl -X POST (https://datacatalog..nx1cloud.com/api/v2/entity?action=ingest \ -H "Content-Type: application/json" \ -d '{ "entityType": "dataset", "aspectName": "datasetProperties", "aspect": { "description": "Updated description" } }' ``` Replace `client` with your client's name. The previous command does the following: 1. Targets a dataset entity 2. Updates the `datasetProperties` aspect 3. Replaces the description with the provided text ### GraphQL API The GraphQL API provides a flexible interface for querying metadata and performing fine-grained updates. Use cases include: * Fetching lineage, schema, or ownership programmatically * Adding tags or ownership to datasets * Automating glossary term assignments For example, you might use the GraphQL API to fetch dataset profiles for an Iceberg dataset. The following API request and response, are an example of this. **API request**: ```bash theme={null} curl -X POST https://datacatalog..nx1cloud.com/api/v2/graphql \ -H "Content-Type: application/json" \ -d '{ "operationName": "getDataProfiles", "variables": { "urn": "urn:li:dataset:(urn:li:dataPlatform:iceberg,retail_banking.completed_accounts,PROD)" }, "query": "query getDataProfiles($urn: String!, $limit: Int, $startTime: Long, $endTime: Long, $filters: FilterInput) { dataset(urn: $urn) { urn type datasetProfiles( limit: $limit startTimeMillis: $startTime endTimeMillis: $endTime filter: $filters ) { rowCount columnCount sizeInBytes timestampMillis partitionSpec { type partition timePartition { startTimeMillis durationMillis __typename } __typename } fieldProfiles { fieldPath uniqueCount uniqueProportion nullCount nullProportion min max mean median stdev sampleValues quantiles { quantile value __typename } distinctValueFrequencies { value frequency __typename } __typename } __typename } __typename } }" }' ``` Replace `client` with your client's name. The previous command does the following: 1. Requests data profiles for the dataset `retail_banking.completed_accounts` in the `PROD` environment 2. Attempts to retrieve: row count, column count, dataset size, and more **API response**: ```json theme={null} { "data": { "dataset": { "urn": "urn:li:dataset:(urn:li:dataPlatform:iceberg,retail_banking.completed_accounts,PROD)", "type": "DATASET", "datasetProfiles": [], "__typename": "Dataset" } }, "extensions": {} } ``` ### When to use APIs Use API-based updates when you are trying to achieve the following: * Integrate DataHub with external systems * Automate updates via pipelines using Airflow, Jenkins, or GitHub Actions * Enforce metadata standards programmatically * Bulk update metadata at scale However, you might prefer to manage metadata in the UI. ## Data quality DataHub provides capabilities for capturing, monitoring, and visualizing data quality rules and test results across datasets. These rules ensure that you can trust the data consumed and quickly identify issues affecting downstream products, models, or dashboards. ### Data quality details When viewing the data quality of a dataset, DataHub displays the following: * List of all assertions/tests containing automated rules that validate the correctness of data * Test results containing pass/fail status with run timestamps * Timestamp of the latest execution * Column-level and table-level checks * Integrated external testing tools such as custom Spark jobs, or checks in Airflow jobs * Associated tags Data quality details You can click a specific assertion to view the following: * Full assertion definition * Historical pass/fail graph * Execution logs or failure summaries Details about a successful passing assertion Sometimes, assertions fail. Details about a failing assertion ### Data quality ingestion Data quality metadata is typically ingested through scheduled pipelines. Supported integrations in NX1 include: * Custom Spark or SQL scripts sending results through the API * Airflow DAGs producing test assertions * Other integrations that can create assertions and send the results using the DataHub API ### Troubleshooting data quality issues When a test is failing, perform the following sequence of actions: 1. Search for the dataset name. 2. Open the dataset overview page and select the **Quality** tab. 3. Identify the failed assertion. 4. Review the failure summary and logs. 5. Use the **Lineage** tab to identify the root cause of the issue Ask yourself, "Is it an upstream table, or an upstream job?." 6. Contact the dataset owners or pipeline owners. 7. Address the issue by either fixing the schema, repairing upstream data, or adjusting the transformation logic. 8. Re-run the test and confirm that it passes. ## Additional resources * To learn about best practices when using DataHub, refer to the [DataHub best practices](./datahub-best-practices) page. * For more details about DataHub, refer to the [DataHub](https://docs.datahub.com/docs/features) official documentation. * If you are using the NexusOne portal and want to learn how to launch DataHub, refer to the [Govern](/documentation/catalog-and-models/tasks/launch-datahub) page.