Using Oracle Big Data SQL to add Hadoop + NoSQL to an Oracle Data Warehouse

As presented at the Dublin HUG IRL meetup, September 2016

1. Mark Rittman, Oracle ACE Director USING ORACLE BIG DATA SQL TO ADD HADOOP + NOSQL TO AN ORACLE DATA WAREHOUSE HADOOP USER GROUP IRELAND (HUG IRL) Dublin, September 2016
2. •Mark Rittman, Co-Founder of Rittman Mead •Oracle ACE Director, specialising in Oracle BI&DW •14 Years Experience with Oracle Technology •Regular columnist for Oracle Magazine •Author of two Oracle Press Oracle BI books •Oracle Business Intelligence Developers Guide •Oracle Exalytics Revealed •Writer for Rittman Mead Blog :
 http://www.rittmanmead.com/blog •Email : mark.rittman@rittmanmead.com •Twitter : @markrittman About the Speaker 2
3. •It’s gone from experimental tech to everyday platform solution •Even Ryanair use it - famously… Everyone’s Talking About Big Data 3 chema-on-read vs schema on write real-time data ingestion agile data provisioning vs. curated data combining Hadoop, NoSQL and Oracle nnel marketing machine learning & decision engines attitudinal vs behavioural data
4. Hadoop Data Lakes Now Standard Part of DW Design 4
5. •Oracle Big Data Appliance - Engineered System for running Hadoop alongside Exadata •Oracle Big Data Connectors - Utility from Oracle for feeding Hadoop data into Oracle •Oracle Data Integrator EE Big Data Option - Add Spark, Pig data transforms to Oracle ODI •Oracle BI Enterprise Edition - can connect to Hive, Impala for federated queries •Oracle Big Data Discovery - data wrangling + visualization tool for Hadoop data reservoirs •Oracle Big Data SQL - extend Oracle SQL 
 language + processing to Hadoop Oracle Software Initiatives around Big Data 5
6. SQL ON HADOOP? 6
7. ISN’T THAT MISSING THE POINT OF HADOOP? 7
8. •Hadoop is not a cheap substitute for enterprise DW platforms - don’t use it like this •But adding SQL processing and abstraction can help in many scenarios: •Query access to data stored in Hadoop as an archive •Aggregating, sorting, filtering and transforming data •Set-based transformation capabilities for other frameworks (e.g. Spark) •Ad-hoc analysis and data discovery in-real time •Providing tabular abstractions over complex datatypes Why SQL on Hadoop? 8 SQL! Though 
 SQL
 isn’t actually
 relational According to Chris Date
 SQL is just
 mappings Tedd Codd
 used 
 Predicate
 Calculus and there’s
 never been
 a mainstream
 relational
 DBMS but it is the
 standard
 language for
 RDBMSs and it’s great
 for set-based
 transforms
 & queries so
 Yes SQL!
9. •Original developed at Facebook, now foundational within the Hadoop project •Allows users to query Hadoop data using SQL-like language •Tabular metadata layer that overlays files, can interpret semi-structured data (e.g. JSON) •Generates MapReduce code to return required data •Extensible through SerDes and Storage Handlers •JDBC and ODBC drivers for most platforms/tools •Perfect for set-based access + batch ETL work But Haven’t We Got That Covered by Apache Hive? 9
10. •Apache Drill is another SQL-on-Hadoop project that focus on schema-free data discovery •Inspired by Google Dremel, innovation is querying raw data with schema optional •Automatically infers and detects schema from semi-structured datasets and NoSQL DBs •Join across different silos of data e.g. JSON records, Hive tables and HBase database •Aimed at different use-cases than Hive - 
 low-latency queries, discovery 
 (think Endeca vs OBIEE) And Isn’t the Future of SQL-on-Hadoop Apache Drill? 10
11. SOMETIMES THOUGH… 11
13. YOU NEED THE ORIGINAL AND BEST 13
15. •Originally part of Oracle Big Data 4.0 (BDA-only) •Also required Oracle Database 12c, Oracle Exadata Database Machine •Extends Oracle Data Dictionary to cover Hive •Extends Oracle SQL and SmartScan to Hadoop •Extends Oracle Security Model over Hadoop •Fine-grained access control •Data redaction, data masking ‣Uses fast c-based readers where possible
 (vs. Hive MapReduce generation) ‣Map Hadoop parallelism to Oracle PQ ‣Big Data SQL engine works on top of YARN •Like Spark, Tez, MR2 Oracle Big Data SQL 15 Exadata
 Storage Servers Hadoop
 Cluster Exadata Database
 Server Oracle Big
 Data SQL SQL Queries SmartScan SmartScan
16. •As with other next-gen SQL access layers, uses common Hive metastore table metadata •leverages Hadoop standard APIs for HDFS file access, metadata integration etc Leverages Hive Metastore and Hadoop file access APIs 16
17. •Brings query-offloading features of Exadata
 to Oracle Big Data Appliance •Query across both Oracle and Hadoop sources •Intelligent query optimisation applies SmartScan
 close to ALL data •Use same SQL dialect across both sources •Apply same security rules, policies, 
 user access rights across both sources Extending SmartScan, and Oracle SQL, Across All Data 17
18. •Read data from HDFS Data Node •Direct-path reads •C-based readers when possible •Use native Hadoop classes otherwise
 
 •Translate bytes to Oracle
 
 •Apply SmartScan to Oracle bytes •Apply filters •Project columns •Parse JSON/XML •Score models How Big Data SQL Accesses Hadoop (HDFS) Data 18 Disks% Data$Node$ Big$Data$SQL$Server$ External$Table$Services$ Smart$Scan$ RecordReader% SerDe% 10110010% 10110010% 10110010% 1% 2% 3% 1 2 3
19. •Oracle Database 12c 12.1.0.2.0 with Big Data SQL option can view Hive table metadata •Linked by Exadata configuration steps to one or more BDA clusters •DBA_HIVE_TABLES and USER_HIVE_TABLES exposes Hive metadata •Oracle SQL*Developer 4.0.3, with Cloudera Hive drivers, can connect to Hive metastore View Hive Table Metadata in the Oracle Data Dictionary 19 SQL> col database_name for a30 SQL> col table_name for a30 SQL> select database_name, table_name 2 from dba_hive_tables; DATABASE_NAME TABLE_NAME ------------------------------ ------------------------------ default access_per_post default access_per_post_categories default access_per_post_full default apachelog default categories default countries default cust default hive_raw_apache_access_log
20. •Big Data SQL accesses Hive tables through external table mechanism •ORACLE_HIVE external table type imports Hive metastore metadata •ORACLE_HDFS requires metadata to be specified •Access parameters cluster and tablename specify Hive table source and BDA cluster Hive Access through Oracle External Tables + Hive Driver 20 CREATE TABLE access_per_post_categories( hostname varchar2(100), request_date varchar2(100), post_id varchar2(10), title varchar2(200), author varchar2(100), category varchar2(100), ip_integer number) organization external (type oracle_hive default directory default_dir access parameters(com.oracle.bigdata.tablename=default.access_per_post_categories));
21. •Run normal Oracle SQL from the Oracle Database server •Big Data SQL query franchising then uses agents on Hadoop nodes to query and return data independent of YARN scheduling; Oracle Database combines and returns full results Running Oracle SQL on Hadoop Data Nodes 21 SELECT w.sess_id,w.cust_id,c.name FROM web_logs w, customers c WHERE w.source_country = ‘Brazil’ AND c.customer_id = w.cust_id
22. •Not all functions can be offloaded to Hadoop tier •Even for non-offloadable operations Big Data SQL will perform column pruning and datatype conversion (which saves a lot of resources) •Other operations (non-offloadable) will be done on the database side •Requires Oracle Database 12.1.0.2 + patchset, and per-disk licensing for Big Data SQL •You need and Oracle Big Data Appliance, and Oracle Exadata, to use Big Data SQL Restrictions when using Oracle Big Data SQL 22 SELECT NAME FROM v$sqlfn_metadata WHERE offloadable ='YES'
23. •From Big Data SQL 3.0, commodity hardware can be used instead of BDA and Exadata •Oracle Database 12.1.0.2 on x86_64 with Jan/Apr Proactive Bundle Patches •Cloudera CDH 5.5 or Hortonworks HDP 2.3 on RHEL/OEL6 •See MOS Doc ID 2119369.1 - note cannot mix Engineered/Non-Engineered platforms Running Big Data SQL on Commodity Hardware 23
24. •No functional differences when running Big Data SQL on commodity hardware •External table capability lives with the database, and the performance functionality with the BDS cell software. •All BDS features (SmartScan, offloading, storage indexes etc still available) •But hardware can be a factor now, as we’re pushing processing down and data up the wire •1GB ethernet can be too slow, 10Gb is a minimum (i.e. no InfiniBand) •If you run on an undersized system you may see bottlenecks on the DB side. Big Data SQL on Commodity Hardware Considerations 24
25. BUT HADOOP IS MORE THAN 
 SIMPLE HDFS FILES 25
27. •Subsequent releases of Big Data SQL have extended its Hadoop capabilties •Support for Hive storage handlers (HBase, MongoDB etc) •Hive partition elimination •Better, more efficient access to Hadoop data •Storage Indexes •Predicate Push-Down for Parquet, ORC, HBase, Oracle NoSQL •Bloom Filters •Coming with Oracle Database 12.2 •Big Data-aware optimizer •Dense Bloom Filters •Oracle managed Big Data partitions Going beyond Fast Unified Query Access to HDFS Data 27
28. •Hive Storage handlers give Hive the ability to access data from non-HDFS sources •MongoDB, HBase, Oracle NoSQL database •Run HiveQL queries against NoSQL DBs •From BDS1.1, Hive storage handlers can be used with Big Data SQL •Only MongoDB, HBase and NoSQL currently “supported” •Others should work but not tested Big Data SQL and Hive Storage Handlers 28
29. •Create Hive table over HBase database as normal •Typically done to add INSERT and DELETE capabilities to Hive, for DW dimension ETL •Create Oracle external table as normal, using ORACLE_HIVE driver Use of Hive Storage Handlers Transparent to BDS 29 CREATE EXTERNAL TABLE tablename colname coltype[, colname coltype,...] ROW FORMAT SERDE 'org.apache.hadoop.hive.hbase.HBaseSerDe' STORED BY 'org.apache.hadoop.hive.hbase.HBaseStorageHandler' WITH SERDEPROPERTIES ( 'serialization.format'='1', 'hbase.columns.mapping'=':key,value:key,value: CREATE TABLE tablename(colname colType[, colname colType...]) ORGANIZATION EXTERNAL (TYPE ORACLE_HIVE DEFAULT DIRECTORY DEFAULT_DIR ACCESS PARAMETERS (access parameters) ) REJECT LIMIT UNLIMITED;
30. •From Big Data SQL 2.0, Storage Indexes are automatically created in Big Data SQL agents •Check index before reading blocks – Skip unnecessary I/Os •An average of 65% faster than BDS 1.x •Up to 100x faster for highly selective queries •Columns in SQL mapped to fields in 
 HDFS file via External Table Definitions •Min / max value is recorded for 
 each HDFS Block in a storage index Big Data SQL Storage Indexes 30
31. •Hadoop supports predicate push-down through several mechanisms (filetypes, Hive partition pruning etc) •Original BDS 1.0 supported Hive predicate push-down as part of SmartScan •BDS 3.0 extends this by pushing SARGable (Search ARGument ABLE) predicates •Into Parquet and ORCFile to reduce I/O when 
 reading files from disk •Into HBAse and Oracle NoSQL database 
 to drive subscans of data from remote DB •Oracle Database 12.2 will add more optimisations •Columnar-caching •Big Data-Aware Query Optimizer, •Managed Hadoop partitions •Dense Bloom Filters Extending Predicate Push-Down Beyond Hive 31
32. •Typically a one-way street - queries run in Hadoop but results delivered through Oracle •What if you want to load data into Hadoop, update data, do Hadoop>Hadoop transforms? •Still requires formal Hive metadata, whereas direction is towards Drill & schema-free queries •What if you have other RDBMSs as well as Oracle RDBMS? •Trend is towards moving all high-end analytic workloads into Hadoop - BDS is Oracle-only •Requires Oracle 12c database, no 11g support •And cost … BDS is $3k/Hadoop disk drive •Can cost more than an Oracle BDA •High-end, high-cost Oracle-centric solution •of course! … So What’s the Catch? 32
33. •Rich, comprehensive SQL access to all enterprise data •Extend Oracle security, advanced analytic features and metadata across Hadoop & NoSQL Oracle Big Data SQL Vision : Unified Query 33
34. Mark Rittman, Oracle ACE Director USING ORACLE BIG DATA SQL TO ADD HADOOP + NOSQL TO AN ORACLE DATA WAREHOUSE HADOOP USER GROUP IRELAND (HUG IRL) Dublin, September 2016