EventDB

EventDB

by on ‎05-27-2015 02:44 PM - edited on ‎10-29-2018 04:13 PM by (3,132 Views)

EventDB takes the data from your filtered streams and stores it into a Postgres-like database using Amazon Redshift. You can then analyze the data using business intelligence tools like Tableau or query the data directly with your preferred SQL client tool.

In this article:

Table of Contents Placeholder

How it Works

Once EventDB is turned on, the prettified view of the schema contains the titles you apply to the filtered streams in AudienceStream, as follows:

  1. The Tealium Collect tag (or other Collect library) sends event data to the Universal Data Hub where it's captured in Tealium's Amazon S3 bucket. At this stage the data is flattened JSON.
  2. Once the S3 bucket reaches a size of 40 MB (uncompressed) or after 1 hour has elapsed, whichever comes first, the data is compressed and prepared for Redshift.
  3. Once the data is compressed, it is copied and imported into the Redshift database for your account.
    The database organization is aligned to your account and streams:
    • A database for each account/profile i.e. ACCOUNT__PROFILE
    • Two tables for each Filtered Stream
      • Raw data attribute column names (ACCOUNT__PROFILE.events_STREAM)
      • Universal Data Hub attribute column names (ACCOUNT__PROFILE.events_view_STREAM)

Turn On EventDB

You must contact your Account Manager to enable EventDB for the appropriate profile(s) in your account.

Once it is activated, it may take up to an hour for data to begin populating your database.

Send Data to EventDB

Use the following steps to send data to EventDB:

  1. Go to Live Events.
  2. Click the pencil icon to edit the stream of data to send to EventDB.

    add_edit_stream.png

  3. In the Edit Stream dialog, slide the Event Data Storage setting to EventDB.
    EventDB Slider with Simplify on.png3
  4. Click Save.
  5. Click Save/Publish to save and publish your changes.

Configure EventDB Attributes

By default, your streams will contain all of the Event Attributes defined in your account. For streams collecting data from the Tealium Collect tag, additional Boolean attributes will be included for each tag that fires during an event. For example, if you have Google Analytics in your iQ account along with the Collect tag, your event streams will include a Boolean attribute named "Google Analytics" indicating whether or not it fired for each event.

Use the following steps to adjust which Event Attributes are saved to EventDB:

  1. From the DataAccess console select the DB tab.
  2. For DB is enabled, click Show EventDB Attributes.

    Attributes marked as Restricted Data are excluded by default but you may check the box to include them. DOM attributes (e.g. URL, domain, referrer, and user agent) are always sent and cannot be excluded.

  3. Check or uncheck the attributes to add or remove from EventDB.
  4. Click Save.
    select data sources.png

Connect to Redshift

To access to your EventDB data, a third-party tool with the ability to connect to a Postgres-style database is required.

  • First Time Users
    The first time you enable EventDB, Redshift Spectrum is automatically enabled.
  • Existing Customers
    Before enabling Spectrum for existing customers that have been using DataAccess, a data migration is needed to ensure all existing and new data are written to the correct location. Coordinate with your team in advance to proceed.
    Learn more about retrieving your credentials from the DataAccess console screen.

Browse the Redshift Database

With your credentials in hand, it's time to connect to the database. The following example uses the freeware program SQL Workbench/J (see Connecting to EventDB with SQL Workbench).

Raw Table Data

There is a raw data table view and a view that joins all the related tables and columns into one convenient location. Notice the difference in column names and that the schema naming convention is [account]__[profile]. The column names are in the same positions in each table, regardless of the view. The main difference is the readability of the entry.

Prettified View of Data