cancel
Showing results for 
Search instead for 
Did you mean: 
Data Engineering
Join discussions on data engineering best practices, architectures, and optimization strategies within the Databricks Community. Exchange insights and solutions with fellow data engineers.
cancel
Showing results for 
Search instead for 
Did you mean: 

Error reading in Parquet file

BL
New Contributor III

I am trying to read a .parqest file from a ADLS gen2 location in azure databricks . But facing the below error:

spark.read.parquet("abfss://............/..._2023-01-14T08:01:29.8549884Z.parquet")

org.apache.spark.SparkException: Job aborted due to stage failure: Task 0 in stage 0.0 failed 4 times, most recent failure: Lost task 0.3 in stage 0.0 (TID 3) (10.139.64.6 executor 0): org.apache.spark.SparkException: Exception thrown in awaitResult:

I searched in google ( as per suggestion in some posts tried to set spark.driver.maxResultSize to 20g , some blogs says to put inferSchema option ) but getting the same error again and again . The file size I am trying to read is 12kb .

I tried with below runtime versions in my databricks cluster

11.3 LTS (includes Apache Spark 3.3.0, Scala 2.12)

11.1 (includes Apache Spark 3.3.0, Scala 2.12)

10.4 LTS (includes Apache Spark 3.2.1, Scala 2.12)

Can anyone please advise how to overcome this issue ?

4 REPLIES 4

anpa
New Contributor III

This error may be related to credential issue.

You can try this code

spark.conf.set("fs.azure.account.auth.type.<storage-account-name>.dfs.core.windows.net", "<your-access-key>")
spark.read.parquet("abfss://............/..._2023-01-14T08:01:29.8549884Z.parquet")

To hide <your-access-key> you can create Secret scopes follow the instructions link below:

Link to create Secret scopes.

BL
New Contributor III

Thanks for your answer .

But I was using same kind of code with access key

BL
New Contributor III

I tried again , but the same error

spark.conf.set("fs.azure.account.key.<ContainerName>.dfs.core.windows.net",ACCESS_KEY)

spark.read.parquet("abfss://............/..._2023-01-14T08:01:29.8549884Z.parquet")

jose_gonzalez
Moderator
Moderator

Can you access the executor logs? When you cluster is up and running, you can access the executor's logs. For example, the error shows:

org.apache.spark.SparkException: Job aborted due to stage failure: Task 0 in stage 0.0 failed 4 times, most recent failure: Lost task 0.3 in stage 0.0 (TID 3) (10.139.64.6 executor 0): org.apache.spark.SparkException: Exception thrown in awaitResult:

Go to the Executor 0 and check why it failed

Connect with Databricks Users in Your Area

Join a Regional User Group to connect with local Databricks users. Events will be happening in your city, and you won’t want to miss the chance to attend and share knowledge.

If there isn’t a group near you, start one and help create a community that brings people together.

Request a New Group