Page 86 - Multicloud Workshop - Prework

Different Structures

In Commercial IT we have structured data and query and save

to a database. When we want to tackle bigdata projects,

rather than centralising the data we spread it out. In this

illustration, we have a client with a query. The data is spread

across a number of nodes and the name node manages the

spread of the data and data locations. The master node deals

with the query and submits jobs to each slave node to query

their local dataset. The master node collates the results and

returns the answer to the client.

© 2016 Engage ESM All Rights Reserved