Running your first Databricks Kensu program

4min

👨‍💻 Modify Notebook to use Kensu
1️⃣ Init Kensu in your code
Add a new code cell at the beginning of the Notebook
Scala
|
%scala
io.kensu.sparkcollector.environments.DatabricksCollector.track(spark, "/dbfs/FileStore/conf/conf.ini" )
%scala
io.kensu.sparkcollector.environments.DatabricksCollector.track(spark, "/dbfs/FileStore/conf/conf.ini" )
﻿
ℹ️  Customizing configuration for single notebook
﻿
It is possible to overwrite the settings of the conf.ini  file only for this particular notebook (i.e. conf.ini could be shared globally, and customizations applied for certain notebook when needed) with the following code:
﻿
io.kensu.sparkcollector.environments.DatabricksCollector.track(spark, "/dbfs/FileStore/conf/conf.ini", "report_to_file" -> "True")
 2️⃣ Run the application
Execute the already existing program. You can find a test notebook here﻿
3️⃣ Check the result
Now you can view the data in the Kensu App.
Initialization with Python code
In case Scala language is not convenient, it is possible to use full Python code
Python
|
def ref_scala_object(jvm, object_name):
    class_name_parts = (object_name + "$.MODULE$").split(".")
    o = jvm
    for p in class_name_parts:
      o = getattr(o, p)
    return o
kensuInit = ref_scala_object(spark.sparkContext._jvm, "io.kensu.sparkcollector.environments.DatabricksCollector")
t2 = spark.sparkContext._jvm.scala.Tuple2
properties = spark.sparkContext._jvm.scala.collection.mutable.HashSet()
#Optional: properties definition
properties.add(t2("offline_file_name", "demo_notebook_test.jsonl"))
properties.add(t2("report_to_file", "True"))
kensuInit.track(spark._jsparkSession, "/dbfs/FileStore/Kensu/conf-1.ini", properties.toSeq())
def ref_scala_object(jvm, object_name):
    class_name_parts = (object_name + "$.MODULE$").split(".")
    o = jvm
    for p in class_name_parts:
      o = getattr(o, p)
    return o
kensuInit = ref_scala_object(spark.sparkContext._jvm, "io.kensu.sparkcollector.environments.DatabricksCollector")
t2 = spark.sparkContext._jvm.scala.Tuple2
properties = spark.sparkContext._jvm.scala.collection.mutable.HashSet()
#Optional: properties definition
properties.add(t2("offline_file_name", "demo_notebook_test.jsonl"))
properties.add(t2("report_to_file", "True"))
kensuInit.track(spark._jsparkSession, "/dbfs/FileStore/Kensu/conf-1.ini", properties.toSeq())
﻿
﻿

Updated 03 Mar 2023

Did this page help you?

TABLE OF CONTENTS

👨‍💻 Modify Notebook to use Kensu

Initialization with Python code

Documentation

Running your first Databricks Kensu program

👨‍💻 Modify Notebook to use Kensu

Initialization with Python code