- Taschenbuch: 626 Seiten
- Verlag: O'Reilly & Associates; Auflage: 2 (3. November 2010)
- Sprache: Englisch
- ISBN-10: 1449389732
- ISBN-13: 978-1449389734
- Größe und/oder Gewicht: 17,8 x 3,8 x 23,3 cm
- Durchschnittliche Kundenbewertung: 1 Kundenrezension
- Amazon Bestseller-Rang: Nr. 343.575 in Fremdsprachige Bücher (Siehe Top 100 in Fremdsprachige Bücher)
Hadoop: The Definitive Guide (Englisch) Taschenbuch – 3. November 2010
Kunden, die diesen Artikel angesehen haben, haben auch angesehen
Es wird kein Kindle Gerät benötigt. Laden Sie eine der kostenlosen Kindle Apps herunter und beginnen Sie, Kindle-Bücher auf Ihrem Smartphone, Tablet und Computer zu lesen.
Geben Sie Ihre Mobiltelefonnummer ein, um die kostenfreie App zu beziehen.
Wenn Sie dieses Produkt verkaufen, möchten Sie über Seller Support Updates vorschlagen?
Über den Autor und weitere Mitwirkende
Tom White has been an Apache Hadoop committer since February 2007, and is a member of the Apache Software Foundation. He works for Cloudera, a company set up to offer Hadoop support and training. Previously he was as an independent Hadoop consultant, working with companies to set up, use, and extend Hadoop. He has written numerous articles for O'Reilly, java.net and IBM's developerWorks, and has spoken at several conferences, including at ApacheCon 2008 on Hadoop. Tom has a Bachelor's degree in Mathematics from the University of Cambridge and a Master's in Philosophy of Science from the University of Leeds, UK.
Was "Java ist auch eine Insel" für die allgemeine Java-Programmierung ist, ist diese Buch für Hadoop. Darüber hinaus werden auch noch optionale Tools und Addons wie Hive, Pig, Snoop usw. behandelt.
Die hilfreichsten Kundenrezensionen auf Amazon.com
I would suggest one thing while reading this book, please have the computer and some IDE open while reading it otherwise you will get lost in theory (which is great nonetheless) and when you actually try to code later, things will go off the top of your head.
Another thing is that, there are various things which you'll understand only after reaching at 4th or 5th chapter. If you are learning Hadoop (from scratch) and you have read its first few chapters only, you may find it difficult but once you reach at 4th or 5th chapters then I'm sure you will find this book amazing.
Then when we started learning from this book, we were able to understand the concepts quite vividly in the beginning 2 chapters yet we were crawling when we reached chapter 4 of Hadoop:Definitive Guide. We got really frustrated and stopped reading this book and decided not to continue it again. But later realizing that it is the very foundation of Hadoop we had to move on.So left with no other option we started with a different plan. This time we started with HIVE.It was quite an easy chapter to our surprise.Then we went on to HBASE. It was PIG that surprised us the most.Even though written by the same author who wrote Chapter 4, this one was pretty simple and illustrative.Now when we found that we were able to proceed through the chapters, we came back to Chapter 5 and then covered the rest of Map-Reduce.But still Chapter 4 is a Mystery....Had to skip it forever...But we found yahoo material explaining serialization pretty well...But couldn't deal with AVRO...Still searching for materials to learn that....:)
I appreciate that this book covers high-level concepts as well as dives deep into the technical details that you will need to know for the design, implementation and day-to-day running of Hadoop and its various associated technologies.
Cloudera CCD-410 certification ranges between tough to very tough. Period.
TRAINING : You are not mandated to take a training. I took a relatively inexpensive training ($300) from edureka dot in, an online training website in India. They give a good overview at 10,000 feet are very good for the price,but no where close enough to get certified. Check out their first session available for free at Youtube. They do have steps to install your own VM, simple project , HIVE,PIG etc. If time and money permits, I strongly suggest going to official cloudera training. It costs about $3000 and includes a free test voucher , so effectively about $2700. Saves you months in preparation time and distinct advantage over your peers that should pay for itself.
Install VM, try few commands, PIG, hive commands, Also try Amazon elastic mapreduce which reduces lot of manual typing and allows you to focus on the coding itself.
LEARNING FROM THIS BOOK: After a training, start with this book. The first Eight chapters are critical (Approximately 300 out of 550 pages). If you are smart,sharp and young , expect to read these eight chapters about three times, more is just fine. Add some time to read rest of chapters once Or twice before the test and all the external links. If you are a busy professional, give a six month window to take the test. Knowing Java is a definitive plus. Buy the Cloudera mock examination after getting comfortable and familiar with Mapreduce($125). It is a nice resource. Explains every answer, links to where you can get more information . Just as an FYI, the real test was far more complex and difficult.
SCENARIOS BASED ON A MAPREDUCE CODE:
You will need to go through the example code, understand what each line does, why it is there, what happens if you comment out a line of the code. As an example,
return job.waitForCompletion(false) ? 0 : -1;
> What does waitForCompletion mean?,
> Is Reduce Job Must Or Optional ?
> How Many Files will running a Map job produce?
> Will the code compile or will it error at run time based on datatypes.?
> What will happen if you run the same job twice ?
> What happens to the map data after the job?
> How does Hadoop handle huge files that cross block boundaries ?
> What happens if you do not explicitly set a mapper or reducer ?
> Will a combiner help , based on a scenario ?
> Which daemon decides the number of Map job to run ?
> How does hadoop handle the blocks when a node crashes?
SCENARIOS BASED ON HIVEQL:
This is an extension of previous scenarios. A small table, a simple SQL query ( example : select stationid,max(temp) from tableX. Answer choice are four set of mapreduce code and you have to chose the right one. Expect to read and understand the mapreduce that emulates how you create a distinct, how you do a sum, average, max, min etc. According to Cloudera website, these are the percentage of questions.
CHAPTER 3 : 17 Percent
CHAPTER 4 : 6 Percent
CHAPTER 5 : 7 Percent
CHAPTER 6 : 18 Percent
CHAPTER 7 : 6 Percent
CHAPTER 8 : 7 Percent
PIG /HIVE/SQOOP/Zookeeper : 8 percent combined (no Hbase)
Chapter no 2 has no reference but is very important. Expect several questions from that chapter since it gives a good overview. Remaining is all the links that cloudera suggests to read and get familier. SQOOP import syntax, creating a hive table via sqoop , creating and populating hive table via sqoop are must knows.
WHY GETTING CERTIFIED:
I have heard the tiring argument that certification is purely academic. Tell that to your doctor or your Dentist. Sound fundamentals are the foundations behind real world experience. Big Data is no different. Understanding the basics will give the confidence; experience will follow while you keep your client happy.
WHY BIG DATA :
My interest on Big Data was spooked by the Harvard Business Review Article claiming that "Data Scientist" was the hottest job of the 21st century. Follow that by googling for "Rayid Ghani", claimed as the data scientist behind Obama's second term victory.
hbr dot org forwardslash 2012 forwardslash 10 forwardslash data-scientist-the-sexiest-job-of-the-21st-century forwardslash ar forwardslash1
OTHER CHOICES :
> Coursera provides a free course "Introduction To Data Science". I signed up for their first batch but could not finish with office commitments.
> Youtube for "Stanford University Hadoop" by Amr Awadallah
I was impressed with these books; You also might like them.
> Big Data: A Revolution That Will Transform How We Live, Work and Think
> Big Data at Work: Dispelling the Myths, Uncovering the Opportunities
> Data Science for Business: What you need to know about data mining and data-analytic thinking
Some day Big Data will become a commodity skillset,but not now. I did a search in glassdoor to see the demand for Hadoop vs some other hot ones. Hadoop is head and shoulders above the rest.
Hadoop - 30,011 postings on Apr 2014
Oracle DBA - 9227 postings ( A Perpetual hot skillset)
Salesforce - 9968 postings
Please post any questions in the comment section and I will certainly try to answer them.
Ähnliche Artikel finden