hadoop - When is mapreduce not suitable for a business? -
Hi, I'm having trouble with areas where mapreads will not be suitable. I understand that there is no point in using Mapreditch with small amounts of data. But what kind of 'complex' questions are not suitable for Maprades?
With data metepadet's petabytes like a business, it would be a good idea to ask questions that say that a type of product is sold right within a month?
But then the complex questions that are generally applicable to a business, do not add any real benefit to it?
Usually everything in the mpadued / hive will be an overall query, but you have non-all queries It could be just a query that could not "perform" a low operation.
The hive can do a fairly complex query by using several queries and windowing functions. Therefore, it is not certain that the statement "Matrudus is suitable with simple totals" questions are completely true.
The types of business questions for the mpadued / hive are not justified, there are real time questions. For example, traffic queries like Top hash tags for Twitter, etc. Overhead will disable them
Or if data is to be normalized for some reason, they should be basically in a table for the mappedad / hive. For example, if you had a high generalized "point of sale" database and would like to query any type of thing that would be painful, unless you first denied the data.
Comments
Post a Comment