Public DBFS root is disabled in Databricks free edition

[Loinguyen318]

I am using notebook to execute a sample spark to write delta table in dbfs using free edition. However, I face an issue, that I can not access the public DBFS after the code executed.

The spark code such as:

data = spark.range(0, 5)
data.write.format("delta").save("/tmp/delta-table")

This is the log error:

Public DBFS root is disabled. Access is denied on path: /tmp/delta-table/_delta_log

[vivadiva]

I am having the same issue. 

I wish to create a new folder

dbutils.fs.mkdirs("/sampledir")

and I get this error

ExecutionError: Public DBFS root is disabled. Access is denied on path: /sampledir

[Sharanya13 · Accepted solution]

Can you use UCVolumes instead of DBFS?

Can you use UCVolumes instead of DBFS? Databricks will disable DBFS as it moves to a serverless approach. I would use UCVoulmes - convenient and governed by UC.Databricks will disable DBFS, as it adopts a serverless approach. I would use UCVoulmes - convenient and governed by UC

[pepco]

Hello,

I would like to re-open this topic if possible. While using UC volumes was marked as accepted answer I would like to ask what are my options. I'm playing with custom logging handler and I would like to write logs into the files. Since DBFS is not enabled and UC volumes don't support direct I/O do I have some other options? 

I could add handler to log into the table I'm not sure if I want to do row-by-row inserts. 

Are there some other options?

[mukul1409]

Yes ,  use UCVolumes instead of DBFS. As Databricks moves toward a serverless architecture, DBFS access is being increasingly restricted and is not intended for long term or production usage. UC Volumes are a better choice than DBFS.