Delta Lake rollback

Question

Need an elegant way to rollback Delta Lake to a previous version.

My current approach is listed below:

import io.delta.tables._

val deltaTable = DeltaTable.forPath(spark, testFolder)

spark.read.format("delta")
  .option("versionAsOf", 0)
  .load(testFolder)
  .write
  .mode("overwrite")
  .format("delta")
  .save(testFolder)

This is ugly though, as the whole data set need to be rewritten. It seems that some meta update would be sufficient and no data I/O should be necessary. Anyone knows a better approach for this?

Answer 1

As of Delta Lake 0.7.0, you can rollback to an earlier version of your Delta Lake table using the RESTORE command. This is a much simpler way to use time travel to roll back your tables.

Scala:

import io.delta.tables._

val deltaTable = DeltaTable.forPath(spark, "/path/to/delta-table")

deltaTable.restoreToVersion(0)

Python:

from delta.tables import *

deltaTable = DeltaTable.forPath(spark, "/path/to/delta-table")

deltaTable.restoreToVersion(0)

SQL:

RESTORE TABLE delta.`/path/to/delta-table` TO VERSION AS OF 0

You can also use the restoreToTimestamp command if you'd prefer to do things that way instead. Read the documentation for more details.