Hadoop s3 integration for Spark

["Afshin, Bardia" <Ba...@capitalone.com>]

Hello community.

I’m considering consuming s3 objects via Hadoop via s3a protocol. The main purpose of this is to utilize Spark to access s3, and it seems like the only formal protocol / integration for doing so is Hadoop. The process that I am implementing is rather formal and straight forward. It will download the contents of a s3 objects, remove some columns from the csv file, and PUT the object into another bucket on s3. Is there any reason doing a simple GET on the object is not as performant if not better than utilizing Hadoop s3a protocol? This is the page that I’m getting my reference from https://wiki.apache.org/hadoop/AmazonS3

________________________________________________________

The information contained in this e-mail is confidential and/or proprietary to Capital One and/or its affiliates and may only be used solely in performance of work or services for Capital One. The information transmitted herewith is intended only for use by the individual or entity to which it is addressed. If the reader of this message is not the intended recipient, you are hereby notified that any review, retransmission, dissemination, distribution, copying or other use of, or taking of any action in reliance upon this information is strictly prohibited. If you have received this communication in error, please contact the sender and delete the material from your computer.

Re: Hadoop s3 integration for Spark

[Steve Loughran <st...@hortonworks.com>]

On 14 Apr 2017, at 17:46, Afshin, Bardia <Ba...@capitalone.com>> wrote:

Hello community.

I’m considering consuming s3 objects via Hadoop via s3a protocol. The main purpose of this is to utilize Spark to access s3, and it seems like the only formal protocol / integration for doing so is Hadoop. The process that I am implementing is rather formal and straight forward. It will download the contents of a s3 objects, remove some columns from the csv file, and PUT the object into another bucket on s3. Is there any reason doing a simple GET on the object is not as performant if not better than utilizing Hadoop s3a protocol? This is the page that I’m getting my reference from https://wiki.apache.org/hadoop/AmazonS3

1. use S3A.  https://www.slideshare.net/steve_l/apache-spark-and-object-stores-for-london-spark-user-group
2. Use Hadoop 2.8 for maximum performance.
3. consider using alternative formats than CSV for your data, as follow-on processing suffers. Better use a column format like ORC and make effective use of predicate pushdowns.


________________________________________________________

The information contained in this e-mail is confidential and/or proprietary to Capital One and/or its affiliates and may only be used solely in performance of work or services for Capital One. The information transmitted herewith is intended only for use by the individual or entity to which it is addressed. If the reader of this message is not the intended recipient, you are hereby notified that any review, retransmission, dissemination, distribution, copying or other use of, or taking of any action in reliance upon this information is strictly prohibited. If you have received this communication in error, please contact the sender and delete the material from your computer.