Athena (serverless)

1. Intro

• run adhoc SQL query serverless, at scale
• Most results are delivered within seconds.
• has integration with AWS glue bts.
• athena call glue data crawler
• crawler, crawls over above source/s and prepare
  • metadata
  • Glue Data Catalog
• Athena uses these while performing query.
• query result will go to:
• S3
• Amazon QuickSight (dashboard)
• 

---

2. pricing

• pay for data Scan : $5 / TB
• save cost
• compressed data
• columnar data
• S3:object.csv --> aws-glue --> parquet(columnar)
• organize data in S3 like /year/month/day/hr/... then scan by date.

---

3. Data Source

• can load data from below source into athena
• Athena will scan,query, analyze data using SQL

3.1 S3

• S3 object:csv,json,avro,parquet(columnar format),
• vpc-logs
• elb-logs
• cloudtrail
• ...

3.2 database

• on-prem/aws (relational/NoSQL) --> aws-glue --> parquet

3.3 KDS - kineses data steam

3.4 DataSource Connector

• backed by AWS Lambda
• to run Federated Queries on RDS,CW,DynamoDB,etc to load data into athena.
• 

---

99. hands on

- bucket-1 >  enable s3-access log.
- copy location-1

athena:
    - create data_source-1
    - Query editor
        - run : create DataBase db1;
        - get query from internet, to DDL/DML to get se-access-log into tables.
        - edit query with location-1.
        - run query to perform ddl and dml, table created : bucket-1_logs
        - check L select * from database-1.bucket-1_logs limit 10.
        - query more standard sql and see report.

 more:
 - recent queries
 - saved queries - also encrypted result.   

99. exam

1. athena read dataset from encrypted s3 (kms-sse). options:
2. share private key with athena
3. kms-sse --> change to sse-s3
4. disable encyption

5. athena automatically decrypts, no need to provide key info