Vertica Integration with H2O Flow: Connection Guide

About Vertica Connection Guides

Vertica connection guides provide basic instructions for connecting a third-party partner product to Vertica. Connection guides are based on our testing with specific versions of Vertica and the partner product.

Vertica and H2O Flow: Latest Versions Tested

Software Version
Partner Product

H2O 3.42.0.3

Partner Product Platform

Red Hat Enterprise Linux release 8.8

Vertica Client

Vertica JDBC 23.3

Vertica Server Vertica Database 23.3

H2O Flow Overview

H2O Flow is the web-based user interface to H2O, an open-source machine learning and artificial intelligence platform.

With H2O Flow, you can capture, rerun, annotate, present, and share your H2O workflows. You can interactively import files and build and tune models. Based on your models, you can make predictions and add rich text to create vignettes of your work.

Download the Vertica Client Driver

H2O Flow uses the Vertica JDBC driver to connect to Vertica. Follow these steps to download the driver:

  1. Navigate to the Client Drivers page on the Vertica website.
  2. Download the JDBC driver package.

    Note  

    For details about client and server compatibility, see Client Driver and Server Version Compatibility in the Vertica documentation.

Download and Start H2O Flow

Note  

H2O requires Java version 7 or later. If you do not already have Java installed on your computer, you can download it from the Oracle website for Java.

  1. On the H2O website, click Download H2O to download the H2O zip file.
  2. Extract the contents of the zip file:

    cd ~/Downloads
    unzip h2o-3.42.0.3.zip
    cd h2o-3.42.0.3
  3. Copy the Vertica JDBC jar file to ~/Downloads/h2o-3.42.0.3 and start H2O Flow:

    java -cp h2o.jar:vertica-jdbc-23.3.0-0.jar water.H2OApp
  4. Open H2O Flow in your browser:

    http://<IP_Address>:54321

Connect H2O Flow to Vertica

  1. In H2O Flow, select ImportSQLTable to import a SQL table into H2O.

  2. Provide the connection URL, user name, and password to connect with Vertica. Provide the name of the table from which you want to import data.

    Click Import to load the data.

    The progress of the data load is displayed:

  3. When the import is successful, click View to view the data frame in H2O Flow.

    The data frame is displayed:

Known Limitations

  • INTERVAL, TIMESTAMPTZ, UUID, BINARY, LONGVARBINARY, and VARBINARY data types are not displayed. A warning message is returned and the data is loaded into H2O as null.
  • For TIMETZ data type, timezone is not displayed.
  • For NUMERIC data type, up to 16 digits to the left of the decimal point and 4 digits to the right of the decimal point are rounded off.
  • For TIME, DATE, TIMESTAMP, and TIMETZ data types, EPOCH time stamp is displayed in milliseconds.

For More Information