Introducing the Oracle "converged" database - Sean Stacey Outbound Product Manager Oracle Server Technologies Feb 9, 2021 ...
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
Introducing the Introducing the Oracle Oracle Converged Database “converged” database Sean Stacey Outbound Product Manager Oracle Server Technologies Feb 9, 2021
Our mission is to help people see data in new ways, discover insights, unlock endless possibilities.
Database Releases and Support Timelines Release andand Database Release Support Timelines Support Timelines 11.2.0.4 12.1.0.2 12.2.0.1 18c Current long term release 19c Current innovation release 21c Premier Support Waived Extended Support Fee Paid Extended Support Limited Error Correction Market Driven Support See Oracle Support Note 742060.1 for the latest schedule. 3 Copyright © 2021, Oracle and/or its a!liates 3
The Future is Data Driven Creating value from data will increasingly determine competitiveness To become Data Driven, Enterprises need Data Driven Apps and Paradigms 5 Copyright © 2021, Oracle and/or its affiliates
Data Driven Apps Create Value from Data in New Ways Machine Real-Time IoT Graph Learning Analytics Analysis Documents Text Spatial Blockchain Search Processing 6 Copyright © 2021, Oracle and/or its affiliates
Data Driven Apps are Built Using New Dev Paradigms Microservices Events SaaS Distributed Data API CI/CD API Driven Defense Low Code Development in Depth 7 Copyright © 2021, Oracle and/or its affiliates
Single purpose databases - advocated by cloud vendors and startups Amazon and other database vendors want customers to run multiple Single-Purpose databases for each • Use of single-purpose databases has been the data type and workload source of a large increase in data management complexity • Require apps to use proprietary APIs for each database • Separately implement security policies in every Amazon database Amazon Amazon Aurora Neptune Timestream • Unique high availability and scalability tools for Amazon DynamoDB each database type Amazon RedShift • Complex management & integration Amazon DocumentDB Amazon Quantum Ledger Database Amazon • End-to-End security, availability, scalability, ElastiCache consistency, etc. is by the weakest of the databases that are used Transac5ons Document and NoSQL Data Warehouse And other single purpose databases 8 Copyright © 2021, Oracle and/or its affiliates
Integration Required Serious integration work required to make a single-purpose database approach feasible on a large scale 9 Copyright © 2021, Oracle and/or its a!liates
In Contrast | Database Offerings Amazon & Niche Database Vendors Oracle Database Amazon Amazon Amazon Amazon Aurora DocumentDB Redshift Quantum Ledger Doc Spatial IoT Blockchain Azure MongoDB Snow!ake Google Synapse Document Store Data Warehouse Big Query Text Reporting ML Graph Analytics Search 10 Copyright © 2021, Oracle and/or its affiliates
In Contrast | Database O!erings Amazon & Niche Database Vendors Oracle Database Amazon Amazon Amazon Amazon Aurora DocumentDB Redshift Quantum Ledger Azure MongoDB Snowflake Google Synapse Document Store Data Warehouse Big Query Analytics • Run single-purpose proprietary database for • Run converged, open, SQL Database for each data type and workload multiple data types and workloads 11 Copyright © 2021, Oracle and/or its affiliates
Unleashing New Capabilities Included in Current Licenses New Functions and Capabilities Modernize Analytics ta 30+ In-database ML IoT Streaming x a dadata F REE E Algorithms w/ 50+ Graph EAnalysis Transactional Event E FR E Algorithms FR E Queues E Datawarehouse Analytics Document store- E FR E FREJSON and XML E and In-memory FRErow ve l ) Blockchain table EE Algorithms FRand column se l e (baformats E FR E Low Code Application Development 100+ Spatial operators E and F R E Improve E!ciency functions E ) FRE PDatabases Container DBS Converged Database p to 3 (u
Oracle Database - A Converged Database Multi-model Key Benefits for Developers, • Relational, JSON, XML, graph, Analysts, and Ops spatial, OLAP • Open Standards SQL across all data Multi-workload plus other popular APIs like REST • Transactions, analytics, in-memory, • Transactional support across all data IoT, streaming, blockchain • Consistent, queryable view of all data • Reliability, scalability, security Multi-tenant • Containers, isolation, orchestration, scale-out with transactional consistency Copyright © 2021, Oracle and/or its affiliates
Oracle Database – Converged Features • Multitenant for Efficient, Agile Database Clouds • AutoML for simple integrated Machine Learning Multitenant In-Memory Hyperscale • Native JSON for Document Data Analytics • Blockchain Tables for Preventing Fraud {} • In-Memory for Database Acceleration • In-Memory Ingest for Fastest IoT JSON In-Memory Cloud IoT Integration • Persistent Memory Store for Lowest Latency • Spatial and Graph for Mapping and Social Networks • Cloud SQL for integrating Object Store Data Lake Blockchain Persistent Machine • Sharding for Hyperscale and Geo Distribution Memory Learning • APEX for Low-Code Development • And many more … Spatial Graph Low Code 14 Copyright © 2021, Oracle and/or its affiliates
Simplify Microservice Architectures with Container Databases • Oracle makes it simple for each Microservice to store its data in a logically separate Data Container aka Pluggable Database • Pluggable Databases can easily be physically combined to simplify deployment, or separated to improve isolation and scalability Microservice App Containers on Kubernetes Pluggable Databases Physical Database Physical Database 15 Copyright © 2021, Oracle and/or its affiliates
DB Nest • Provides OS Level Isolation for both Container and Pluggable Databases from other databases and applications running on CDB NEST the same server PDB 1 PDB 2 PDB 3 NEST • Utilizes Linux Namespaces to provide OS level isolation NEST NEST • Each CDB resides in a single parent nest • PDBs reside within individual child nests created within the parent. • Each PDB Nest has it’s own PID and cannot access other PDBs in other nests OPERATING SYSTEM • Enabled with DBNEST_ENABLE set to CDB_RESOURCE_PDB_ALL 16 Copyright © 2021, Oracle and/or its affiliates
Oracle In-Database Machine Learning Bring Algorithms to the Data, Not Data to Algorithms • Oracle DB natively includes dozens of ML algorithms including deep learning • Perform parallel ML directly in Data Warehouse or OLTP for fast model building and real time scoring of new data • Keep data secure by avoiding copy contagion – copies of data in analytics systems are a common source of data breaches • Extend in-database algorithms with Python and R • Easily deploy models via REST API • Monitor model performance to recommend model rebuild 18 Copyright © 2021, Oracle and/or its affiliates
AutoML in OML4Py Automatically build and compare Machine Learning models Auto Model Auto Feature Auto Tune Selection Selection Much faster than >50% reduction Significant score exhaustive search in features improvement Data ML Model • Auto Model Selection • Auto Feature Selection • Auto Tune Hyperparameters – Identify in-database algorithm – Reduce # of features by – Significantly improve that achieves highest model identifying most predictive model accuracy quality – Improve performance and – Avoid manual or exhaustive search – Find best model faster than with accuracy techniques exhaustive search Enables non-expert users to leverage Machine Learning 19 Copyright © 2021, Oracle and/or its a!liates
AutoML in OML4Py Automatically build and compare Machine Learning models = Apply ML Model Data Be!er Predictions • Auto Model Selection • Auto Feature Selection • Auto Tune Hyperparameters – Identify in-database algorithm – Reduce # of features by – Significantly improve that achieves highest model identifying most predictive model accuracy quality – Improve performance and – Avoid manual or exhaustive search – Find best model faster than with accuracy techniques exhaustive search Enables non-expert users to leverage Machine Learning 20 Copyright © 2021, Oracle and/or its affiliates
Native JSON Database • Oracle makes it easy to use the best format for each use case • Oracle Database natively supports JSON and XML • Index any JSON element for fast OLTP access • Makes it simple to discover the structure of any JSON Document • Trivially create relational views across JSON collections • ACID and Parallel SQL Analytics across ALL formats 23 Copyright © 2021, Oracle and/or its affiliates
Improved JSON Speed and Flexibility • New high performance binary JSON • SQL UPDATE now allows SET and REMOVE datatype in SQL and PL/SQL of JSON elements • Scans up to 10x faster • Declarative transformations for JSON • Updates up to 4x faster CREATE TABLE j_order UPDATE j_order ( SET po_doc = JSON_TRANSFORM( po_doc, id INTEGER PRIMARY KEY, SET '$.address.city' = 'Kona', po_doc JSON REMOVE'$.phones[*]?(@.type = "office")' ); ) WHERE id = 555; 24 Copyright © 2021, Oracle and/or its a!liates
Introducing Blockchain Tables • Blockchain technology enables records to be linked together using a cryptographic hash • Makes it simple to detect and prevent fraud LEDGERS_OF_TRADE • This technology has been complicated for business to adopt ID User Value Hash 1 Tom 500 ADSJS 2 Carol 176 %SHS 3 Steve 500 SH@1 4 John 176 DHD3 • Introduced in Oracle Database 20c 5 Mike 332 *EGG • Simple to use and integrate into existing applications 6 Sarah 632 AH11 7 Eve 25 LIO$ • Part of the standard database o!ering 8 Prisha 850 SHS4 BLOCKCHAIN TABLE 29 Copyright © 2021, Oracle and/or its a"liates
Blockchain Tables Secure ledger table managed by a trusted provider to prevent fraud • Specialized table that allows normal SQL inserts and queries • CREATE BLOCKCHAIN TABLE LEDGER_OF_TRADES • NO DROP UNTIL 31 DAYS IDLE • NO DELETE LOCKED • Rows are cryptographically chained, chain is verifiable by participants • Blockchain table can participate in queries with other tables • Orders of magnitude easier to use and more functional than existing blockchains 30 Copyright © 2021, Oracle and/or its affiliates
Demo
Database In-Memory In-Memory In-Memory In-Memory Self-Managing Spatial and Text Vector Joins Hybrid Scans In-Memory In-Memory In-Memory In-Memory Row Store Column Store Vector Join Column Store (Buffer Cache) Column Store Resume Experience Education SALES ITEMS Relational Spatial Text 10x faster spatial analytics 5-10x faster joins 10x faster hybrid queries Auto in-memory 3x faster text analytics using SIMD vector populate / evict instructions No need to declare tables INMEMORY 32 Copyright © 2021, Oracle and/or its affiliates
Self Managing In-Memory Column Store • When INMEMORY_AUTOMATIC_LEVEL is set to HIGH Oracle automatically manages the placement and removal of objects in the In-Memory Column Store • Works on objects that have no pre existing In-Memory settings IN-MEMORY COLUMN STORE • Tracks usage patterns and moves and evicts objects from the column store • Columns are automatically compressed based on their usage patterns Simplifies use and improves efficiency COLUMNAR DATA MAY BE COMPRESSED AUTOMATICALLY INACTIVE OBJECTS REMOVED FROM COLUMN STORE ACTIVE DATA IS AUTOMATICALLY CONVERTED TO COLUMNAR FORM AND MOVED TO COLUMN STORE 33 Copyright © 2021, Oracle and/or its affiliates
Database In-Memory – Base Level • Customers can now create an In-Memory column store of upto 16GB without having to license the Database In-Memory option • Not all In-Memory features are available with Base-level • Also made available in 19.8 Release Update • Enabled by setting parameter: INMEMORY_FORCE = BASE_LEVEL • Must be configured at CDB level 34 Copyright © 2021, Oracle and/or its affiliates
Streaming Insert for IoT • New in-memory insert algorithm • Declare table MEMOPTIMIZE FOR WRITE • Client performs low-latency insert into in-memory buffer • High throughput bulk insert done asynchronously in background • 25 million inserts per second on two socket server 35 Copyright © 2021, Oracle and/or its affiliates
SQL Macros SQL> Scalar Macros provide a simple way to encapsulate complex SQL expressions • Acts like a pre-processor for SQL. Reusable code transparent to SQL Optimizer. • Makes it easy to create reusable and portable code • Simplifies the calling of common SQL Expressions • No expensive context switching create or replace function SQL> select sales_tax(20,'BEER') from dual; sales_tax(unit_cost number, unit_type varchar) SALES_TAX(20, 'BEER') return varchar2 SQL_MACRO(SCALAR) is -------------------- begin 24 return q'[case when unit_type = 'FOOD' EQUIVALENT T0 then unit_cost else unit_cost * 1.2 end]'; SQL> select case end; when 'BEER' = 'FOOD' then 20 FUNCTION RETURNS STRING else 20*1.2 end from dual; 36 Copyright © 2021, Oracle and/or its affiliates
SQL Macros for Tables SQL> SQL Table Macro encapsulate the SQL used in the FROM clause • Used in ! Parametrized Views ! Polymorphic Tables CREATE or replace FUNCTION orders_waiting_to_ship(order_value number) RETURN varchar2 SQL_MACRO(TABLE) SQL> select * IS BEGIN from RETURN q'[ orders_waiting_to_ship(200); SELECT i.* FROM ORDERS o, ORDER_ITEMS i where order_status > 6 and o.order_total >= order_value and o.ORDER_ID = i.ORDER_ID ]'; end orders_waiting_to_ship; 37 Copyright © 2021, Oracle and/or its affiliates
Persistent Memory Store • Native Persistent Memory Store keeps database data and redo in local Persistent Memory (PMEM) – For workloads that need lower latency than flash memory • SQL runs directly on data stored in mapped Persistent DATA GUARD Memory file system – Eliminates IO code path, no need for large buffer cache – New database algorithms prevent partial or inconsistent stores to Persistent Memory • Requires Data Guard for protection from server failures or persistent memory failures • Scale-out using Native Database Sharding 38 Copyright © 2021, Oracle and/or its affiliates
Oracle Graph Scalable, 50 built-in functions, query language, developer APIs Description • For highly connected datasets • Relationships are !rst class citizens o i.e. they are represented directly • Specialized graph languages, like SPARQL or Gremlin, e"ciently query & navigate highly connected datasets Advantages • Ability to make frequent schema changes • Quickly make relationships between di#erent types of data • Real-time query response time Use Cases • Fraud detection, social networking, recommendation engines, knowledge graphs, data lineage 42 Copyright © 2021, Oracle and/or its affiliates
Spatial data Oracle Spatial used by nearly every major government globally Description • Optimized for storing & querying data that represents objects defined in geometric space o Define points, lines & polygons o Compute length, area, distance o Query relationships between geometries o e.g. “Is there a residence located within a mile of the area we are planning to build the landfill?” Advantages • Efficiently processes spatial data types • Manages geographic & location-data in native type • Supports wide range of applications Use Cases • Automated mapping, facilities management, GIS, wireless location services, location-enabled business 45 Copyright © 2021, Oracle and/or its a!liates
Cloud SQL - Query of Cloud Object Stores • Enables high-performance SQL over • Oracle Object Store • AWS S3 • Azure Blob Storage • Supports many file formats • Including text, CSV, JSON, Parquet, ORC and Avro • Scalable joins with data lakes simplifies • Loading data into data warehouse • Connecting to data lake sources 47 Copyright © 2021, Oracle and/or its affiliates
Oracle APEX • Industry leading low code app dev platform • Easy to learn and get started • SQL and REST friendly • Build responsive & mobile apps • Included (free) with all Oracle Databases • Part of Always-Free Oracle Cloud • Global success stories across all industries • Loved by 500K+ developer community 20x faster with 100x less code Used to build Oracle COVID-19 Therapeutic Learning System (https://covid19.oracle.com) 48 Copyright © 2021, Oracle and/or its affiliates
Oracle APEX 20.1 New Features Friendly URLs APEX apps now use SEO-friendly URLs at runtime. Mega Menus Application navigation menus can be rendered as a Mega Menu, an expandable menu that displays all menu entries in a single floating panel. Remote Deployment Provides a one-click deployment option to promote an APEX app to another system, all from within the web UI of APEX. Automatic Backups Provide the ability for applications to be automatically backed up as part of daily maintenance, or on-demand. Export as Zip Files Export a component-level, hierarchical representation of the app in zip format. Suitable for deploying granular application changes and maintaining within version control systems. Data Loading for Oracle Load data directly from OCI Object Store, as well as S3. Support added for Avro and Parquet file Autonomous Database formats. Service Level for Oracle Developers can now specify the performance level of queries in APEX apps within Autonomous Autonomous Database Database. Redwood UI The user interface of APEX and the App Builder has been refreshed to Redwood, Oracle's new user experience design language. 49 Copyright © 2021, Oracle and/or its a!liates
What about Security? • Innovations with password management • In built Compliance • Uni!ed Auditing additions • Encryption Keys and Key management • Key Vault multi-master clustering 50 Copyright © 2021, Oracle and/or its affiliates
New | Free Autonomous Database Runs on same infrastructure as ADW/ATP Always-free “micro” instances with up to - 2 always free ADB instances per OCI tenancy - 20 GB database storage per instance - 1 OCPU per instance Full functionality, full experience, including - APEX - REST Interface (ORDS) - SQL Developer Web - Machine Learning Notebooks 56 Copyright © 2021, Oracle and/or its affiliates
Database Features & Licensing App Part of Oracle Documentation Interactively compare releases/features h!ps://pex.oracle.com/database-features 57 Copyright © 2021, Oracle and/or its affiliates
Oracle Cloud Infrastructure New Free Tier Always Free Services you can use for unlimited time + oracle.com/cloud/free 30-Day Free Trial Get $500 in free credits
Contact Me: • Sean.Stacey@oracle.com @SeanStaceyfwiw In/sean-stacey-oracle seanstacey.org Sign up for your Always Free cloud account: h!ps://www.oracle.com/cloud/free Oracle Database Features and Licensing: h!ps://apex.oracle.com/database-features/ 59
You can also read