News for hackers. Stuff that matters.

A comment on Command-line tools can be faster than your Hadoop cluster. Parent comment

Re:Command-line tools can be faster than your Hadoop cluster by simonster on Sunday January 18, 2015 @03:05PM (#8908646)
If you want disk parallelism, RAID 0 is probably easier than Hadoop.
[ Parent | on HN ]
Re:Command-line tools can be faster than your Hadoop cluster by cmurf on Sunday January 18, 2015 @03:16PM (#8908686)
That would depend on the data set and the strip size. Striping is good for streaming. Linear/concat of 2+ drives with XFS would be faster with a lot of files than end up in separate AG's on separate drives which can be accessed in parallel.
[ Parent | on HN ]