App Trino

Trino Installation

The Trino Installation capability in xDP allows you to deploy, configure, and manage a production-ready Trino cluster on your compute infrastructure with a streamlined, wizard-driven workflow. It automates the complex setup of the distributed SQL query engine, enabling your analytics teams to run fast, interactive queries across disparate data sources—such as data lakes, warehouses, and relational databases—almost instantly. This accelerates time-to-insight by removing the operational overhead of manual Trino deployment and configuration.

Key Concepts

• Federated Query: The ability to use a single SQL query to access and join data from multiple, distinct data stores. Trino acts as a universal query layer, eliminating the need for complex data movement or ETL pipelines for analytical workloads.
• Data Store Catalog: A Trino configuration that defines how to connect to a specific data source (e.g., an S3 bucket, a MySQL database, or HDFS). xDP automatically generates these catalog configurations from the connections you register in the Data Store module, simplifying integration.
• Trino Coordinator & Workers: The core components of a Trino cluster. The Coordinator is responsible for parsing SQL statements, planning queries, and managing worker nodes. The Workers execute the tasks and process the data. xDP manages the deployment of both components as scalable services on your cluster.
• Metadata Management: Trino requires a relational database (like PostgreSQL) to store its internal metadata. xDP provides the flexibility to use a managed, internal PostgreSQL instance for simplicity or to connect to your own external database for compliance and control.

Capabilities

• Automated Provisioning: Deploy a fully configured Trino cluster, including the coordinator and workers, onto your target compute cluster in minutes.
• Simplified Configuration: Manage Trino settings, resource allocation (CPU/memory), and data source connections through an intuitive Form Editor or a powerful YAML-based interface.
• Seamless Data Store Integration: Automatically create and configure Trino catalogs by selecting from a list of pre-registered xDP Data Stores, ensuring immediate connectivity to your data.
• Lifecycle Management: Easily update configurations, scale resources, or uninstall the Trino application directly from the xDP interface.

Tutorial (Getting Started)

Prerequisites

• You have an active and healthy Compute Cluster registered in xDP.
• You have the necessary permissions to install applications on the target cluster.
• (Recommended) You have configured at least one connection in the Data Store section that you wish to query with Trino.
• Supported version: Trino 476 ,468

Your First Workflow

1. Select the Trino Application: From the xDP sidebar, navigate to Platform > Apps. Ensure your desired compute cluster is selected in the top navigation. Find the Trino card in the application catalog and click Install.
2. Configure the Database Connection: Trino requires a PostgreSQL database for its metadata. For a quick start, leave the Use Internal PostgreSQL option enabled. xDP will automatically provision a dedicated database for this Trino instance. Click Next. Tip If your organization requires using a centrally managed database, you can disable this toggle and provide the connection details for your own external PostgreSQL instance.
3. Select Data Stores: This step connects Trino to your data. If you have pre-configured Data Stores, they will appear here. Select the ones you want to query. If you have none configured yet, you can click Skip Data Stores for now and add them later. Click Continue.
4. Configure Application Settings: You can fine-tune Trino's deployment settings using either the Form Editor for guided inputs or the YAML Editor for advanced control. For this initial installation, the default settings are sufficient.
5. Deploy Trino: Review your settings and click Save and Apply Configuration. xDP begins provisioning the Trino components on your compute cluster.
6. Verify Installation: After a few moments, the Trino application card on the Apps page will show an Installed status, indicating that your Trino cluster is ready for use.

How-to Guides

Connect Trino to a New Data Store After Installation

1. Navigate to Platform > Apps and locate your installed Trino application.
2. Click the Edit button on the Trino application card.
3. In the installation wizard, navigate to the Select Data Stores step.
4. Click Configure Data Stores and select the new data source you wish to add.
5. Proceed through the wizard and click Save and Apply Configuration. xDP will update the Trino cluster with the new catalog configuration.
6. Confirm the application status returns to Installed. You can now query the new data source.

Customize Trino Resource Allocation

If your queries are resource-intensive, you may need to increase the CPU or memory allocated to the Trino coordinator or workers.

1. Navigate to Platform > Apps and click Edit on your Trino application.
2. Go to the Configure Application step in the wizard.
3. Select the YAML Editor tab.
4. Locate the resources section for the component you want to adjust (e.g., trino-coordinator or trino-worker).
5. Modify the limits and requests values for cpu and memory.
6. Click Save and Apply Configuration to redeploy Trino with the new resource settings.
7. Verify the pods restart successfully and the application status is Installed.

Best Practices

• Right-Size Your Cluster: Start with the default resource allocations and monitor query performance and pod resource utilization. Adjust CPU, memory, and worker replica counts based on your workload to optimize for cost and performance.
• Centralize Data Store Connections: Before installing Trino, define all your primary data sources in the xDP Data Store module. This creates a single source of truth for connection configurations and makes integrating them with Trino and other applications trivial.
• Use the Form Editor for Simplicity: For common changes like adjusting replica counts or simple environment variables, the Form Editor provides a safe and user-friendly interface that prevents syntax errors.
• Version Control for YAML: For complex or business-critical deployments, use the YAML editor. Copy the configuration and store it in a version control system like Git to track changes and facilitate disaster recovery.
• Isolate Workloads: Consider deploying separate Trino clusters for different teams or use cases (e.g., ad-hoc analytics vs. production reporting) to prevent resource contention.