13
jhom
6y

When a client refuses to use cloud services for Hadoop in lieu of 4 local machines... For TB scale data.

Comments
  • 2
    Cause 4 smaller VMs shouldn't be used for Terabytes of data, edited to clarify that
  • 4
    Depends on the data. You can built solid storage systems that are going to be less expensive over time. But as I said, it depends on the data, how it's being accessed, written and how fast it has to be
  • 1
    @Kimmax this is for data science and eventually machine learning. With like 48 gb ram and 12 cores (summation of all workers) . It works... Yea. Effectively? Depends on your definition of effective :/
Add Comment