Exam AWS Certified Data Analytics - Specialty topic 1 question 115 discussion

lzo format focuses on high compression & decompression speed. So C & E

C & D . As per the link https://aws.amazon.com/blogs/big-data/top-10-performance-tuning-tips-for-amazon-athena/ preferred compression is Gzip . For Athena, we recommend using either Apache Parquet or Apache ORC, which compress data by default and are splittable. When they are not an option, then try BZip2 or Gzip with an optimal file size.

C & E are correct answers

Generrally, high compression equal low speed decompression. gzip is more high compression than lzo. so I choose E. C and E are correct I think.

C AND D ARE THE CORRECT ANSWERS

Seems no one mentioned that LZO is splitable on text file while gzip is not. When we use parquet format, we intend to utilize its features that it is splitable, compressible. CSV are file is one example of the text file. So it seems LZO should be chosen over Gzip. Refer to https://aws.amazon.com/cn/blogs/big-data/top-10-performance-tuning-tips-for-amazon-athena/

C & D is the correct Answer for me

https://docs.aws.amazon.com/whitepapers/latest/aws-glue-best-practices-build-performant-data-pipeline/building-a-performance-efficient-data-pipeline.html

D and E would not convert the data to a column format, so I am not sure why and how it would help improve the latency, therefore only viable options are B (use Athena to unload parquet files), and C (use glue to convert to parquet and partition the data). Also the questions asks to choose two solutions.

CE: I passed the test

GZip is often a good choice for cold data, which is accessed infrequently. Snappy or LZO are a better choice for hot data, which is accessed frequently.

Answer should be C and E.

Correct answers are C & D as AWS recommends using either Parquet or ORC columnar data stores and BZips or Gzip compression. For Athena, we recommend using either Apache Parquet or Apache ORC, which compress data by default and are splittable. When they are not an option, then try BZip2 or Gzip with an optimal file size. Option A is wrong as MySQL Workbench does not improve query performance. It instead increases operational overhead. Option B is wrong as Athena is not ideal for running batch jobs. Use Glue instead. Option E is wrong as although .lzo is supported and can provide better compression and decompression speeds, Gzip is recommended for Athena as per the AWS documentation.

E over D: https://docs.aws.amazon.com/athena/latest/ug/compression-formats.html It is stated that Lzo compression format has less compression than Gzip, but this makes decompression much faster, which will result in improved query performance and since that's what's needed my answer is E. There is no cost concern in the question so it doesn't matter if files stored in s3 are bigger with Lzo compression.

cd as exp[lained by others

Answer should be B & C

GZIP is default compression format for Parquet