Presto Installation : A Step by Step Guide to Run SQL Queries

    By: 

    Saurabh Mahawar, Developer Relations Engineer at IBM

    Install and run PrestoDB (0.296+) on your local machine in under 10 minutes. Follow this hands-on tutorial to deploy a high-performance SQL query engine for your Data Lakehouse architecture.

    Prerequisites

    Before getting started, ensure that the following are installed:

    • Operating System: macOS (Intel or Apple Silicon), Linux, or Windows.
    • Java Development Kit (JDK): Java 17 is now required for newer Presto versions. (Java 8 is no longer sufficient). (Run java -version to confirm).
    • Python: Version 2.7 or 3.x (Presto’s launcher uses Python). (Run python3 --version to confirm).
    • RAM: Minimum 4GB available (8GB+ recommended).

    Download and Extract

    Run these commands in your terminal to download the latest server package.

    • Open terminal/command line interface and run below command’s.
    # 1. Download the tarball
wget https://repo1.maven.org/maven2/com/facebook/presto/presto-server/0.296/presto-server-0.296.tar.gz

# 2. Extract it
tar -xvzf presto-server-0.296.tar.gz

# 3. Move it to your Desktop or your preferred directory
mv presto-server-0.296 ~/Desktop/presto-server
cd ~/Desktop/presto-server
    • Verify that the resulting directory structure matches the example below:

    Configure the Server

    • Inside presto-server–*, create the etc configuration directory.
    mkdir etc
    • Create the following four files inside the etc directory: config.properties, jvm.config, log.properties, and node.properties.
    • Apply the following configurations to the node.properties file. (This identifies your machine).
    node.environment=production
node.id=ffffffff-ffff-ffff-ffff-ffffffffffff
node.data-dir=/path-to-your-presto-server/data -----> Set this path to your presto server directory location
    • Apply the following configurations to the jvm.config file. (This tunes the Java engine. Use this exact config to support Java 17).
    -server
-Xmx2G
-XX:+UseG1GC
-XX:G1HeapRegionSize=32M
-XX:+UseGCOverheadLimit
-XX:+ExplicitGCInvokesConcurrent
-XX:+HeapDumpOnOutOfMemoryError
-XX:OnOutOfMemoryError=kill -9 %p
--add-opens=java.base/java.io=ALL-UNNAMED
--add-opens=java.base/java.lang=ALL-UNNAMED
--add-opens=java.base/java.lang.reflect=ALL-UNNAMED
--add-opens=java.base/java.lang.invoke=ALL-UNNAMED
--add-opens=java.base/java.util=ALL-UNNAMED
--add-opens=java.base/java.util.concurrent=ALL-UNNAMED
--add-opens=java.base/java.util.concurrent.atomic=ALL-UNNAMED
--add-opens=java.base/java.nio=ALL-UNNAMED
--add-opens=java.base/sun.nio.ch=ALL-UNNAMED
-Djdk.attach.allowAttachSelf=true
    • Apply the following configurations to the config.properties file. (This sets up your node as both a Coordinator and a Worker).
    coordinator=true
node-scheduler.include-coordinator=true
http-server.http.port=8080
query.max-memory=1.5GB
query.max-memory-per-node=1GB
query.max-total-memory-per-node=1.2GB
discovery-server.enabled=true
discovery.uri=http://localhost:8080

      query.max-memory (1.5GB) must be smaller than the JVM Heap (2GB).

      • Apply the following configurations to the log.properties file. (This sets your default logging level).
      com.facebook.presto=INFO

      Add Data Connectors

      • Create the catalog folder inside etc directory.:
      mkdir etc/catalog
      • Create a file named etc/catalog/tpch.properties and apply below configuration:
      connector.name=tpch

      Start the Server

      • Navigate to the presto-server-* directory and run command:
      bin/launcher start
      • Wait 10 seconds, then check the status:
      bin/launcher status
      • Access the Presto UI at http://localhost:8080 via your web browser.

      Run Your First Query through Presto CLI

      • Download the CLI from here and move the jar file to presto-server-* directory with below command:
      mv ~/Downloads/presto-cli-0.296-executable.jar ./presto
      • Make it executable:
      chmod +x presto
      • Connect to your local cluster:
      ./presto --server localhost:8080 --catalog tpch --schema tiny

      Once you see the presto:tiny> prompt, you can type your SQL:

      SHOW TABLES;
      • Run a standard TPC-H aggregation query
      SELECT returnflag, linestatus, sum(quantity) as sum_qty, count(*) as count_order
FROM lineitem
GROUP BY returnflag, linestatus;
      • To stop the server cleanly:
      bin/launcher stop

      Troubleshooting Common Issues

      If the server fails to launch, investigate these three common configuration pitfalls:

      • Port 8080 already in use

      The Error: java.net.BindException: Address already in use The Fix: Another app (often another web server or Jenkins) is using port 8080.

      • Option A: Kill the process using it: lsof -i :8080 then kill -9 <PID>.
      • Option B: Change Presto’s port in etc/config.properties to 8081.
      • Java version not supported

      The Error: Presto requires Java 17+ (found 1.8.0_xxx) The Fix: You are running an old Java version.

      • Run java -version to check.
      • If you have valid Java 17 installed but it’s not default, point to it explicitly in your startup command: JAVA_HOME=/path/to/java17 bin/launcher start
      • Server Starts but Dies Immediately

      The Error: Launcher says “Started” but status says “Not running”. The Fix: This is usually a configuration typo or memory issue.

      • Check the detailed logs: cat data/var/log/server.log
      • Ensure query.max-memory in config.properties is less than -Xmx in jvm.config.

      Follow Us