What is Apache Spark?

Advanced tech...except for audio

I was so frustrated with all videos out there. This one is quite good!

I am sorry for doubting my right earphone.

really great overview of spark, helped me understand what problems it solves, what the other tools in the space are, and why spark is currently preferred

Audio quality is poor making it hard to understand. Just one audio channel.

Dear Cloudera: Please change this title to: What is MapReduce

Mike Olson walks into a hairdresser: "give me the Frankenstein's Monster"

Am I def? My left ear can't hear anything, I know I have problems in that ear, but wow, I can't hear a thing.

So apache spark is written in scala and is meant to be used with hadoop

This isn't an explanation of Spark. More of a cloudera advertisement showing how proud Apache was on building Spark.

The right side of my headphones is broken.

With macOS: System Preferences -> Accessibility -> Audio -> "Play stereo audio as mono".

give me the executive summary! This did NOT help.

Copy the video link, open Vlc , from Media menu : Open network stream, paste the link there. Play the video , from Audio menu : Stereo mode: left.

Bingo !

I see you more like a marketing agent trying to sell his product instead of somebody who tries to enlighten people what it is

I'm sure Spark is much better that this annoying video

being into IT and NOT being able or not being bothered to do some simple audio enhancement is a shame

To summarize, Spark is easier to use than Map Reduce because it is easier to code. Map Reduce is hard because it is hard to code. Spark can be used in all sorts of applications which involve Big Data. The video ends before he could give out anything more than that....

is that good choice to start directly learning spark without​ hadoop

That guy is amazing

Audio quality is poor making it hard to understand..!

This guy seems to be some HR/Marketing guy. There is NO SINGLE WORD what is spark and why it's different from M/R.

problem with mapreduce is that no one wants to write java code (atleast not those who do not have java coding experience). its simply too verbose and looks pre historic. Also, using pig/hive is unreliable. we have done a project where pig is used and the output in final tables is not what was expected. On ther other hand, wirting scala/python code in Spark is really terse ! with few lines you can achieve what you need. it simply makes that map reduce framework look outdated

My right-brain is now better informed

Remove humming noise

Well said..!

Well said!

programming MapReduce requires a lot of Java and understanding of complex api`s and abstractions, Spark is a faster simpler processing engine for working on hadoop data with ability for in memory computation and batch processing organization. simply spark is a newer method for working with hadoop data processing being designed better for this, and works well with big data languages like Python, Scala (works well with Java), yet spark is not the one size fits all or the last of the crop, as demands shift so will an introduction of new working paradigms and software spin offs, and each part of the hadoop ecosystem has its strengths in different data crunching requirements, the good thing is all the data is in one place with hadoop.

One picture is worth a thousand words. No diagrams here. do over

So, What is Apache Spark ?

where is audio????

It is a great Spark presentation if you know about Hadoop.
Well done Cloudera.

Sorry, but this does not explain what Spark really does/is, when to use it etc. If you already know Spark then maybe you think it explains it well. But to someone uninitiated, it adds zero value.