Databricks job cluster setup
You will need to install the library according to the Databricks runtime being used.
| Runtime version | Spark version | Installation method |
|---|---|---|
| 13.3 | Spark 3.4.1 | Upload and install from DBFS |
| 16.4 | Spark 3.5.2 | Upload and install from Workspace |
To install the library on job clusters you can follow these steps:
-
- Upload the PerfGazer JAR to a location accessible by the job cluster (e.g., DBFS).
-
- You can use an init script to install the PerfGazer library on the job cluster, for example:
if [ -f "/databricks/jars/perfgazer_spark_3-5-2_2_12_0_0_29_SNAPSHOT.jar" ]; then
rm -f "/databricks/jars/perfgazer_spark_3-5-2_2_12_0_0_29_SNAPSHOT.jar"
fi
cp -f /dbfs/FileStore/jars/perfgazer_spark_3-5-2_2_12_0_0_29_SNAPSHOT.jar /databricks/jars