Azure Databricks : The backend could not get session tokens for path /mnt. Did you remove the AWS key for the mount point?

Question

I'm trying to Mount Azure Blob storage containers to DBFS. Implementation is as below. got below following error . Not sure why I'm getting an error Did you remove the AWS key for the mount point? when I'm trying to connect to azure blob storage.

Do I missing anything here? Could you please help me solving this issue - Thanks --

Code Implementation

dbutils.fs.mount(
  source = "wasbs://<container-name>@<storage-account-name>.blob.core.windows.net",
  mount_point = "/mnt/iotdata",
  extra_configs = {"fs.azure.account.key.<storage-account-name>.blob.core.windows.net":dbutils.secrets.get(scope = "<scope-name>", key = "<key-name>")})

Error

ExecutionError: An error occurred while calling o275.mount.
: com.databricks.backend.daemon.data.common.InvalidMountException: The backend could not get session tokens for path /mnt. Did you remove the AWS key for the mount point?
    at com.databricks.backend.daemon.data.common.InvalidMountException$.apply(DataMessages.scala:612)
    at com.databricks.backend.daemon.data.filesystem.MountEntryResolver.resolve(MountEntryResolver.scala:84)

Abhishek K · Answer 1 · 2021-09-01T05:32:21.497

I followed below steps to successfully mount and read data in dataframe from Azure blob storage.

Step1: Import pyspark and SparkSession

import pyspark
from pyspark.sql import SparkSession

Step2: Use below command to unmount any existing mount point

dbutils.fs.unmount("/mnt")

Step3: To get key of storage

Go to storage account and select Access keys

Now click on Show keys as shown in below screenshot

Then copy first key.

Step4: To mount blob storage use below code:

dbutils.fs.mount(
  source = "wasbs://<container_name>@<storage_account_name>.blob.core.windows.net",
  mount_point = "/mnt/Sales",
  extra_configs = {"fs.azure.account.key.<storage_account_name>.blob.core.windows.net": "key"})

From your given code you must remove dbutils.secrets.get(scope = "<scope-name>", key = "<key-name>") and replace it with Azure blob storage key only.

Step5: Read data in dataframe from blob storage container

Step6: Dataframe

Step7: Original CSV file in blob storage

it didn't work for me. Are there any permissions I need to look into before mounting? or any extra permissions i need to give on the access key or scope ?? — Sri, Aug 31 '21 at 19:02
No, there are no permissions needed. Steps given in above answer are enough. I have updated my answer to show where you can get key of storage account. — Abhishek K, Sep 01 '21 at 05:33

Azure Databricks : The backend could not get session tokens for path /mnt. Did you remove the AWS key for the mount point?

1 Answers1