Troubleshooting Fire/Databricks Integration

When the workflow is executed, nothing shows up in Fire

One problem might be that the postbackURL is not configured right in Fire Insights under Administration/Configuration.

The other problem can be that the machine running Fire Insights is not accessible from the Databricks Cluster. Test connectivity to the Fire Insights machine from Databricks.

Connectiving from Databricks to Fire postbackURL can be done in Databricks via Notebooks using the telnet command.


When the workflow is executed, nothing shows up in Fire

Another reason might be that you are using the Databricks High Concurrency cluster. Ensure that you are connecting Fire to Databricks Standard cluster.

When accessing most Databricks pages in Fire, it gives Simba JDBC error

The probably reason for it is that the Databricks Simba JDBC jar file is not deployed in Fire.

In the workflow editor, it shows ‘Cannot connect to Fire’

Ensure that under Administration/Configuration, app.runOnCluster is set to false.

Checking the cluster logs in Databricks

There are times when it is helpful to look at the Cluster logs in Databricks when running Fire with Databricks.

The following logs under Driver Logs are useful:

  • log4j-active.log
java.lang.Exception: An error occurred while initializing the REPL. Please check whether there are conflicting Scala libraries or JARs attached to the cluster, such as Scala 2.11 libraries attached to Scala 2.10 cluster (or vice-versa).
at com.databricks.backend.daemon.driver.DatabricksILoop$class.initSpark(DatabricksILoop.scala:98)

This error can happen when running spark 2.3 version of Fire with spark 2.4 cluster on Databricks. Either upgrade Fire to spark 2.4 version, or create another Databricks cluster which supports spark 2.3.

Databricks Cluster Versions Support

Databricks Runtime Version Spark Version Scala Version

6.2 2.4.4 2.11

6.3 2.4.4 2.11

6.4 2.4.5 2.11

6.5 2.4.5 2.11