mapreduce tutorial | what is mapreduce | hadoop tutorial | mapreduce with example | Trendytech
Вставка
- Опубліковано 25 сер 2024
- Want to learn Big Data by Sumit Sir?
Checkout the Big Data course details here: trendytech.in/...
Consistent 5 Star Google Rated Big Data Hadoop & Spark Course
Extended 20 weeks Online Big Data course designed for Working Professionals. This will help you master the concepts in a lot of depth.
A single full fledged 150+ hours Big Data Course with a bunch of optimization techniques specially designed to help you crack top big data companies with ease.
All aspects related to big data Hadoop, scala, spark, Big Data on AWS Cloud are covered in depth.
More emphasis on Hands On learning so that you get really confident.
Live Capstone projects, Regular assignments & assessments.
A Wide range of Interview questions covered along with Resume preparation & career guidance.
Regular Doubt clearing sessions & Dedicated Student mentor support.
Our students works for some of the leading MNCs like walmart, target, brillio, cisco, GE Healthcare, UHC, JP Morgan etc.
Most of our Students get multiple offers in a very short span with hikes ranging from 40% to as much as 250%
If someone do this Big Data Course wholeheartedly, with dedication and zeal to learn, then there is no looking back. It will lead to guaranteed success.
#bigdata #dataengineering
Checkout the Big Data course details here: trendytech.in/?referrer=youtube_bd3
Very well explained sir.... Clearly understood how a mapreduce works.... Thank u
Glad to hear that
awesome explanation sir...
Loved the session 🙌🪄
thanks a lot sir. However one question, when you have 1 separate node for reducer and 4 nodes for mapper, is not the cluster should be called 5 node cluster instead of 4 node cluster ? And in case reducer is one among the 4 nodes, then how overhead of having reducer on one of the existing mapper node gets handled by Map- Reduce or Hadoop ?
How come data from Map phase fit into single reducer machine? There should be 4 reducers right? It would have been better if you would have taken this example with multiple reducer machines. Now it is clear until the map phase but Reducer is still confusing. As if there are multiple reducers and let's say data is skewed then let's say the key value pair ("Hello",1) occurs so many times that it does not fit in one Reducer machine how will be take care of that?
@tonystarc9567
My understanding is following -
- we can have multiple reducer machine and that depends on following -
1- no of available cores
2- shuffled data partition
3- key on which data is partitioned.
If data is skewed and we have only one key then there will be only partition of data , and it will be processed sequentially by one core only. The tasks will be queued up on single core and will be processed one after the other.
is bigdata is in demand now?