hadoop fs -du gives two data columns

Question

I ran the following command on my test hadoop instance :

hadoop fs -du /test/data/

51179082   153537246   /test/data/9875/2016/02/03
46949272   140847816   /test/data/9875/2016/02/04

I understand du gives the size of files and directories, but for hadoop fs -du it is giving two sizes. Can some one help me understand what are these?

Hilliary

its is similar to the unix du , since you have 2 folders , it showed the folder specific size – Shot

I dont think so, because i now have three folders and it is giving 2 columns for each of those – Hilliary

Answer 1

One column is size of the file and the other column is size of the file after replication.

Hilliary

which is which? – Squirm

@Squirm The first is the size and the second is the space consumed, i.e. the post-replication size. So a replication of 3 would cause three times the size to be consumed in space. – Pemphigus

I have a 1 node cluster. The cluster complains of under-replicated blocks. However the 2nd value shows nearly 3x of the 1st value! – Implicative