0

I understood that hdfs snapshot keeps tracks of added or deleted files from a directory. How is the behaviour when i have files (PARQUET) that are appended continuously?

djohon
  • 705
  • 2
  • 10
  • 25

1 Answers1

1

When you create a snapshot of a directory/file, they are added in the subdirectory /.snapshot , so they are ordered by date ascending whatever the file format is! There's no a maximum number of snapshots.

hdfs snapshot keeps tracks of added or deleted files from a directory

Correct me if I'm wrong, but a snapshot keeps track of every single change (even in the file) and not just of the added and deleted files from a directory.

I hope this helps you to understand their behaviour!

HDFS snapshots documentation

hey_rey
  • 103
  • 8