Hi,
I needed to understand one concept on snapshotted files deletion.
I was figuring out the concepts of HDFS snapshot and got stuck in a concept of deletion:
a) If we delete a file, the snapshot for which is created already, and we remove it from .Trash as well. This deletion actually shows that the file is deleted. I compared the disk usage for the folder as well and confirmed that the size got reduced by the amount of the size of the deleted file.
b) However, when I see the files in the .snapshot directory, I could see all of them (even the deleted one). And adding the individual file sizes of snapshots gets the total size before deletion.
Anyone got any ideas what’s happening in the actual sense? I mean I do understand that the file seems to be deleted but is it actually deleted? Because what I get from the theory on Apache about snapshot concept is that it doesn’t actually creates a copy of the file which is snapshotted & I tested it as well. But this deletion concept, I’m stuck.
Thanks in advance,
Adil