What’s Coming in

Oracle Database 23c

Accelerating our mission to simplify development of data-driven apps

Jenny Tsai-Smith
Vice President
Oracle Database Product Management
Safe harbor statement

The following is intended to outline our general product direction. It is intended for
information purposes only, and may not be incorporated into any contract. It is not
a commitment to deliver any material, code, or functionality, and should not be
relied upon in making purchasing decisions. The development, release, timing, and
pricing of any features or functionality described for Oracle’s products may change
and remains at the sole discretion of Oracle Corporation.

Copyright © 2023, Oracle and/or its affiliates

Modern Apps – Born in the Cloud

Modern data-driven apps use a new generation of technologies leveraging the flexibility and efficiency of cloud to
enable more compelling experiences for users and deliver more value from that data

New Types of Data New Types

New Dataof Types
Analytics New Workload Types

{ JSON }

Relational Documents Spatial Text Data Machine Graph Lake House Micro Geo- IoT Blockchain
Warehouse Learning Services Distributed

Copyright © 2023, Oracle and/or its affiliates

Oracle Data Management Vision

Make it simple to develop and run

modern operational and analytic apps
for all use cases, at any scale

Copyright © 2023, Oracle and/or its affiliates

Delivering on the Vision
Complete and Simple data platform for developing and running modern apps and analytics

Complete Simple

Converged Database Autonomous Database

Complete support for all modern data types, Simplest cloud database for developers,
workloads, and development styles DBAs, data analysts
Completely consistent, scalable, available, and Simplest cloud database for running any app at
secure any scale or criticality

Copyright © 2023, Oracle and/or its affiliates

Build a Data-Driven App with a Converged Database Engine
Add a SQL statement (or a REST API call), not another database

Store and Access movie details in Find movies customers have in common Use Fuzzy Text Search to find movie reviews
JSON documents using Graph Analytics containing “disappointed” or variations of it

CREATE TABLE movie_details( SELECT c1, e, m, e1, c2 SELECT title, comments

title VARCHAR2(255), FROM MATCH FROM movie_reviews
movie JSON); (c1)-[e]->(m)<-[e1]-(c2)
WHERE CONTAINS(comment,
on CUST_MOVIE
'fuzzy(disappointed,
WHERE c1.cust_id= 1246813
SELECT m.title Title, 70, 6, weight)', 1) > 0;
AND C2.cust_id= 1002487
m.movie.director DIR,
LIMIT 100;
m.movie.Star STAR
FROM movie_details m;

Find theaters within 5km of Jane’s location Store rental transaction in a Blockchain Store concession purchases in XML and easily
using built-in Spatial functions Table to prevent fraud retrieve them using standard SQL

SELECT theater.name CREATE BLOCKCHAIN TABLE rental( CREATE TABLE purchase_orders (

FROM theater, u_id number, key_column VARCHAR2(10),
customer user_name varchar2(100), xml_column XMLType);
WHERE customer.name = 'Jane' order_date date,……);
AND SDO_WITHIN_DISTANCE(
SELECT xml_column
theater.location,
FROM purchase_orders;
customer.location, INSERT INTO rental VALUES (1,'Dominic',’08-
'distance=5 unit=km') FEB-2023’,..);
='TRUE';
Copyright © 2023, Oracle and/or its affiliates
 Oracle’s Converged Database Is Available Everywhere

Oracle Cloud
Infrastructure

Cloud@Customer 3rd Party

Clouds Same database,
same skills

Commodity
Hardware

Converged
Database
On-Premises Via Containers

Copyright © 2023, Oracle and/or its affiliates

Oracle Autonomous Database is a Fully-Managed Cloud Service
With automation across all components driven by AI and predictive analytics

Exadata Oracle Automated Data Center

Infrastructure & Database Ops and Machine
Automation & Automation Learning

Copyright © 2023, Oracle and/or its affiliates

Simplest Integration with Modern Cloud Ecosystem

Autonomous Database has seamless integration with the full range of cloud services that Apps need
Cloud App Dev Services

DevOps Container Resource API Functions Artifact Streaming

CI/CD Registry Manager Gateway Registry

Cloud Runtime Services

-
Autonomous
Database Kubernetes Vault Identity Web-App Data Analytics Object
Engine (Key Mgmt.) Firewall Science Cloud Storage

Uniquely enables the lowest cost and highest performance cloud applications

Copyright © 2023, Oracle and/or its affiliates

Autonomous Database – Runs Anywhere
Same autonomous database service available in the cloud and on-premises

Oracle Dedicated Region Exadata Hybrid and

Public Cloud Cloud@Customer Cloud@Customer Multicloud
Configurations

Copyright © 2023, Oracle and/or its affiliates

Oracle Autonomous Transaction Processing ranked highest out of 17 vendors
in all four use cases in the 2022 Gartner® Critical Capabilities for Cloud
Database Management Systems for Operational Use Cases.

Source: Gartner Critical Capabilities for Cloud DBMS for Operational Use Cases, Merv Adrian, Rick Greenwald, Henry Cook, Xingyu Gu, December 2022
Gartner is a registered trademark of Gartner, Inc. and/or its affiliates in the U.S. and internationally and is used herein with permission. All rights reserved Gartner does not endorse
any vendor, product or service depicted in its research publications, and does not advise technology users to select only those vendors with the highest ratings or other designation.
Gartner research publications consist of the opinions of Gartner’s research organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed
or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose.

Copyright © 2023, Oracle and/or its affiliates

Oracle Autonomous Data Warehouse ranked second out of 17 vendors in all
four use cases in the 2022 Gartner® Critical Capabilities for Cloud Database
Management Systems for Analytical Use Cases.

Source: Gartner Critical Capabilities for Cloud DBMS for Analytical Use Cases, Rick Greenwald, Merv Adrian, Adam Ronthal, Henry Cook, Philip Russom, Xingyu Gu, December
2022
Gartner is a registered trademark of Gartner, Inc. and/or its affiliates in the U.S. and internationally and is used herein with permission. All rights reserved Gartner does not endorse
any vendor, product or service depicted in its research publications, and does not advise technology users to select only those vendors with the highest ratings or other designation.
Gartner research publications consist of the opinions of Gartner’s research organization and should not be construed as statements of fact. Gartner disclaims all warranties, expressed
or implied, with respect to this research, including any warranties of merchantability or fitness for a particular purpose

Copyright © 2023, Oracle and/or its affiliates

Oracle Database RoadMap

Oracle Database Oracle Database Oracle Database

19c 21c 23c

Long Term Release Innovation Release Long Term Release

2019 2021 2023

Copyright © 2023, Oracle and/or its affiliates

Oracle Database
23 c
Oracle Database 23c accelerates Oracle’s
mission to make it simple to develop and run all
data-driven apps

Over 300 new features in 23c App Simple

BETA available since October 2022, on-
premises and in Oracle Cloud

https://tinyurl.com/OracleBeta
Copyright © 2023, Oracle and/or its affiliates
A Better Document Database Than MongoDB

JSON Data Guide In-Memory

Blockchain Document
JSON
Tables
Oracle Database Analytics

JSON
Schema
23 c MongoDB API Full SQL
Support

JSON / Relational Duality Native JSON

Best in Class Security
Datatype
2X Faster
Operational
Workloads than
Graph Analytics
10X faster
MongoDB JS Stored
Procedures
Analytics

Copyright © 2023, Oracle and/or its affiliates

A Better Operational Graph Database

High Speed Ingest 50+ Graph JSON or Row

SQL/PGQ Oracle Database Analysis
Algorithms
based Graph
Models

23
In-Memory
Analytics
c Flashback
Queries

Graph
Studio Low-Cost
Unrivaled High Secondary
Availability Average Indexes

24/7
of 26X Comprehensive
Faster Best in Class Security
Enterprise
Full SQL
Support Capabilities

Copyright © 2023, Oracle and/or its affiliates

Better SQL Than Anyone Else
Automatic Flashback
Indexing, Partitioning,
Query SQL For
Approximate
Functions
Materialized Views
Oracle Database JSON

Window Functions Real Time

Materialized
Views
23 c Results
Cache

Analytical
Polymorphic Readers don’t Multi Views
Tables
block Writers Model
SQL Model
Clause SQL
Pattern
External SQL/PGQ
Tables
SQL Macros Matching Parallel Query

Copyright © 2023, Oracle and/or its affiliates

Oracle Database 23c Developer Focus Areas
Accelerating our mission simplify development of data-driven apps

Operational In-Database
JSON Relational Duality
Property Graphs Microservices Support

SQL SQL SQL

CACHE CACHE CACHE

App Server App Server App Server

JavaScript SQL Enhancements

True Cache Stored Procedures

Copyright © 2023, Oracle and/or its affiliates

Simple Update or Delete Via a Join

It is complex to write apps that need to Update or

Delete data based on data stored in another table -- Update or delete rows base on foreign keys
-- constraints without a sub-select or IN clause

UPDATE emp e
Now this is simple since Update or Delete statements SET e.sal = e.sal*2
can include joins with other tables FROM department d
WHERE e.deptno = d.deptno
AND d.name = 'Development';

Copyright © 2023, Oracle and/or its affiliates

Update RETURNING Clause Enhancements

It is complex to write apps that need the old and new

value of data that is being updated -- Update stmts can now return new and old val

UPDATE emp e
SET e.sal = e.sal*2
Now this is simple since UPDATE statements can return WHERE e.deptno = 10
the old and new values for a column RETURNING OLD sal, NEW sal
• Also works for MERGE INTO :old_sal, :new_sal;

Copyright © 2023, Oracle and/or its affiliates

GROUP BY Column Alias or Position

It is complex to write SQL statements that group by

and return complicated expressions -- Group by column alias

SELECT substr(state,1,2) as st,

SUM(total) as revenue
Now this is simple, using column aliases inside the FROM sales
GROUP BY clause GROUP BY st;
• It also works with column positions: GROUP BY 2

Copyright © 2023, Oracle and/or its affiliates

SELECT Without FROM Clause

It is unnecessarily complex to write a complete SQL

statement to retrieve the value of an expression from the -- SELECT expressions without FROM Dual clause
database
SELECT sydate;

SELECT 2*2;
Now this is simpler using a SELECT without a FROM
clause SELECT MY_Func();
• No need to SELECT expr FROM DUAL
SELECT dbms_random.value*10
CONNECT BY LEVEL <=10;

Copyright © 2023, Oracle and/or its affiliates

Table Value Constructor

It is complex to run multiple SQL statements to operate

on multiple rows -- New VALUES clause allows for multiple rows in
-- SELECT, INSERT, MERGE statements

INSERT INTO t1(cola, col_b, col_c)

VALUES
Now this is simple because the VALUES expression can
(1, 'string a', sqrt(9)),
construct multiple rows (2, 'string b', power(2,2)),
• Supports INSERT, SELECT, and MERGE (3, 'string c', 1e6);

SELECT *
FROM (VALUES ( 1,'SCOTT' ),
( 2,'SMITH' ),
( 3,'JOHN’ )
) t1 (employee_id, first_name);

Copyright © 2023, Oracle and/or its affiliates

IF [NOT] EXIST DDL Clause

It is complex to determine when there are genuine

errors in scripts that create and drop tables -- Do something if an object does/doesn’t exist.
-- Useful to avoid errors in test scripts
• Because actual errors are masked by false errors that are
returned if tables already exist CREATE TABLE IF NOT EXISTS emp
(EMPNO NUMBER(4) NOT NULL,
ENAME VARCHAR2(10),
JOB VARCHAR2(9),
Now this is simple using IF NOT EXIST syntax …
• Supports CREATE, DROP ALTER DEPTNO NUMBER(2)
);

DROP TABLE IF EXISTS t1;

Copyright © 2023, Oracle and/or its affiliates

Boolean Datatype

It is a complex and manual process to map language

and JSON boolean types to text or number columns in
CREATE TABLE customers(
SQL cust_id number,
Active boolean);

-- Find our active customers

This is now simple using SQL Boolean data type
SELECT cust_id
FROM customers
New SQL syntax WHERE active;
• IS [NOT] TRUE, IS [NOT] FALSE
• WHERE TRUE AND NOT Boolean_column
• SELECT 1 > 2
• etc.

Copyright © 2023, Oracle and/or its affiliates

4096 Column Tables

It is complex to write applications that need more than

1000 columns in a table -- Requires compatibility parameter set to 23
• Machine Learning Models
• Analytics that denormalize many columns ALTER SYSTEM
SET max_columns = EXTENDED;

CREATE TABLE orders(… );

Now this is simple since tables can contain 4096
columns

Copyright © 2023, Oracle and/or its affiliates

Annotations

It is complex to record and share application

metadata -- Annotate the table and columns
CREATE TABLE emp (
For example: id number ANNOTATIONS(Identity,
• Mark data as sensitive (GDPR, PCI, etc.) Display 'emp id'),
ename varchar2(50) ANNOTATIONS(Display 'emp name'),
• Indicate which columns should be hidden in UI sal number ANNOTATIONS(Display 'salary',
UI_hidden)
)
ANNOTATIONS (Display 'Emp Table', App_Version '2.0');
This is now simple using Annotations inside the
database -- Find annotations usage for Emp table and its columns
SELECT *
FROM user_annotations_usage
Annotations can be defined on tables, views, MVs, WHERE Object_Name = 'EMP'
columns, indexes, and domains during creation or AND Object_Type = 'TABLE'
AND Column_Name IS NOT NULL;
via an ALTER statement ;

Copyright © 2023, Oracle and/or its affiliates

SQL Domains

It is complex to write applications that use user-defined

types (object types) in the database
--Create a new domain to identify credit cards
• Converting app-tier language types to database user-
defined types makes programs more complex CREATE DOMAIN credit_card AS varchar2(16);

-- Use the new domain when creating a contacts table

Now Extended ISO SQL Domains provide a simple CREATE TABLE customers(
c_name varchar2(50),
alternative for common cases
cr_card_no DOMAIN credit_card,

…);
Domains specify the intended usage of data
• Like the “format” annotation in OpenAPI
• Examples: credit card, password, email, telephone,
URL, quantity, etc.

Copyright © 2023, Oracle and/or its affiliates

SQL Domains

Applications can query the Domain of a column and

take appropriate action
--Create a credit cards domain and specify
• Hide password, verify a credit card number constraints and display format

CREATE DOMAIN credit_card AS varchar2(16)

Domain definitions can optionally specify data CONSTRAINT
constraints, display format, and collation CHECK(SUBSTR(cr_card,1,2) in
('34','37', '41', …, '54','55'))
Domains are non-disruptive DISPLAY '************'||SUBSTR(cr_card,12,4);

• Adding a domain to a column does not require any

modification to SQL that uses the column

A column's domain can depend on the value of another

column, implementing FlexFields

Copyright © 2023, Oracle and/or its affiliates

Schema privileges

It is complex to manage privileges on all the database

tables, views, and procedures used by an application -- Grant Scott SELECT on all tables in sales schema

GRANT SELECT ANY TABLE

ON SCHEMA sales
Now this is simple using GRANT on a Schema TO scott;
• More secure than granting * ANY privileges to users

Copyright © 2023, Oracle and/or its affiliates

DEVELOPER Role

It is complex to grant all the privileges developers need

to create, debug, and tune apps -- Provide user Scott all privileges need to create new app

GRANT DB_DEVELOPER_ROLE TO scott;

This is now simple using the DEVELOPER role

The DB_DEVELOPER_ROLE includes :

• ADMINISTER SQL TUNING SET
• CREATE ANALYTIC VIEW
• CREATE ATTRIBUTE DIMENSION
• CREATE CUBE
• CREATE CUBE BUILD PROCESS
• Etc.

Copyright © 2023, Oracle and/or its affiliates

Oracle Database 23c Automation, Security
Continuing our mission to make operating and securing Oracle databases simpler

SALES

RESULTS

ORA-600

ML Augmented Real-time Real-time SQL Plan Automatic Materialized

Statistics Management Views

TLS 1.3

SQL Firewall TLS 1.3 Support Azure AD

OAuth2 Integration

Copyright © 2023, Oracle and/or its affiliates

Enhanced Real-time Statistics Using Machine Learning

A new ML model allows Oracle to predict how data will change over time in order to
improve optimizer statistics
Stale statistics are one of the top reasons of bad optimizer plans
• Volatile tables go through heavy and continuous DMLs
• Regular statistics gathering practice cannot keep up with data changes leading to stale statistics

Real-time Statistics uses a regression model and extrapolates the “current” stats
• Stats advisor checks the snapshots daily and marks a table as volatile or not
• Conditions must be met for columns to be used in the model, keeping the impact low and model creation fast

Benefits:
• Reduce impact on operational systems by reducing frequency of statistics collection
• Accurate statistics improve query execution, allowing for businesses to become more responsive to dynamically
changing data

Copyright © 2023, Oracle and/or its affiliates

Real-time SQL Plan Management

SQL Plan Management has been enhanced to detect and prevent SQL performance
regressions in real-time instead of as a background process

Prevents performance regressions

• Plan changes are detected at parse-time
• Performance of previous SQL execution plans are compared with performance of the new plan
• If a previous plan out-performs the new plan, the previous plan will be enforced in future executions of the
SQL statement

Reduces risk by immediately finding, testing, and repairing SQL plans

Improves application availability

• Repairs are automatic, eliminating the need for slow manual intervention

Copyright © 2023, Oracle and/or its affiliates

Automatic Materialized Views

In Oracle Database 23c, Automatic Materialized Views' internal algorithms have been
enhanced to take advantage of a broader range of features in the database
• Automatic Materialized Views can take advantage of Automatic Partitioning
• Enhanced internal costing model for Automatic Material Views selection, balancing:
• Access benefits
• Maintenance (refresh) cost
• Frequency of executions
• Broader rewrite capabilities available, including outer-join queries with filter predicates

Benefits:
• Automatically balances the trade-off between the operational overhead of maintaining materialized
views, and the query performance benefits of materialized views

Copyright © 2023, Oracle and/or its affiliates

In-Database SQL Firewall

Oracle SQL Firewall is an easy-to-use firewall solution, with minimal performance and operation overhead,
for ALL Oracle Database deployments

• SQL Firewall is embedded and built into the database, ensuring that it cannot be bypassed
• Embedding also gives the firewall full visibility into the top-level SQL, stored procedures and the related
database objects

• Oracle SQL Firewall offers real-time protection against common database attacks by monitoring and
blocking “unauthorized SQL” and SQL injection attacks from inside the database
• It first collects all SQL that should be allowed (“allow-list”), and then detects, blocks, and logs any
unexpected SQL
• SQL Firewall can also use session context data such as IP address for its enforcement

Copyright © 2023, Oracle and/or its affiliates

Support for TLS 1.3

Transport Layer Security (TLS) is used to secure data in transit from the database server
• Destinations include database, end-users, applications, and identity providers (Active Directory,
OCI IAM…)
• Approved standard since August 2018
• TLS supports newer security methods than Database Network Native Encryption (NNE)

Starting with Oracle Database 23c, Oracle Database supports TLS version 1.3
• TLS requires a server-side certificate which can be based on a common root certificate, or can be
self-signed
• Improves the performance of TLS handshakes making connections faster
• TLS version 1.3 is the latest and most secure TLS protocol
• Implements encryption, authentication and integrity for data in transit

Copyright © 2023, Oracle and/or its affiliates

Azure AD OAuth2 Integration

Azure AD OAuth2 Integration supports Azure OAuth2 token-based authentication and authorization
for Oracle database access

New open-source plugins for OCI and Azure

• Allows seamless future enhancements to OCI and Azure support as well as support for new
cloud services without having to update the JDBC driver, and without customers having to
recompile code
• Oracle JDBC furnishes simplified configuration for OCI and Azure plugins with a minimal set
of connection properties

Benefits:
• Security: token-based access with token expiration interval allows secure authorization and
authentication.
• Multicloud: the integration with Azure Active Directory simplifies multicloud database access.

Copyright © 2023, Oracle and/or its affiliates

Oracle Database 23c Manageability, Performance, Availability
Continuing our mission to make operating and securing Oracle databases simpler
INVENTORY

ID LO DES QTY
C C
HIGH 101 35 Quantity < 20
MED LOW 10 35
1

Inter-Instance Resource Lock-free Column

Management Priority Transactions Value Reservations

APP SERVER

COORDINATOR

SHARD 1 SHARD 3

RO RO
IRELAND GERMANY USA
SPAIN FRANCE
ITALY

Read-only Per PDB RAC on Kubernetes and

Standbys Podman Sharding Enhancements

Copyright © 2023, Oracle and/or its affiliates

Inter-Instance Resource Management

Server Level Inter-Instance Resource Management enables specifying the minimum CPU
resources needed by each CDB and sharing of CPU resources above the minimum

• Integrates with Linux cGroups to implement CPU shares and CPU limits
• Allows a server to be over-subscribed, yet guarantee CPU resources for a particular
database instance
• If a database instance does not use its guaranteed CPUs at any moment in time, the
other database instances can use it

Copyright © 2023, Oracle and/or its affiliates

Priority Transactions

Priority Transactions automate the process of aborting a low-

priority transaction that holds a row lock that blocks a high-
priority transaction

How it works: ID LOC DESC QTY

• A new parameter sets the priority (HIGH, MEDIUM, LOW) of a user 35

101
transaction
• Users can configure the maximum time a transaction with will wait before
aborting a lower priority transaction holding a row lock

Benefits:
• Reduces the administrative burden for DBAs
• Maintains response time and transaction throughput for high priority
transactions

Copyright © 2023, Oracle and/or its affiliates

Lock-free Column Value WITHOUT LOCK-FREE Reservation
Reservations Long running
transaction blocks

With Lock-free Reservations, multiple ID LOC DESC QTY

other users

transactions can make reservations to the 101 35

UPDATE INVENTORY
SET QUANTITY = QUANTITY -1
same column value WHERE ID = 101

• Example, a shopping app can reserve inventory UPDATE INVENTORY

for an item when the item is put in a shopping SET QUANTITY = QUANTITY -1
WHERE ID = 101
basket without preventing other shoppers from
purchasing the item
• Updates to column values are treated as a lock- WITH LOCK-FREE Reservation Lock-free Reservations allow
free reservation, an intent to change the value in other users to run TX if stock is
available
the future
ID LOC DESC QTY

• Row containing value is locked and updated only

at commit time 101 35 UPDATE INVENTORY
SET QUANTITY = QUANTITY -1
• Database honors all reserved updates until WHERE ID = 101

quantity exceeds a pre-defined limit UPDATE INVENTORY

ALTER TABLE INVENTORY MODIFY
(QUANTITY ESCROW CONSTRAINT SET QUANTITY = QUANTITY -1
QUANT_ESC CHECK(QUANTITY >= 25)); WHERE ID = 101

Copyright © 2023, Oracle and/or its affiliates

Read-only Per PDB Standbys
In Oracle Database 21c, we introduced Per-PDB Data Guard (DGPDB)
RO
• Supports two primary Container DBs running on different sites RO

• Each PDB is open Read-Write on one site at a time

• Protection on PDB or CDB level using real-time Apply
• Each PDB can independently failover to remote site so no need to fail over a full Container Database!
• Per PDB automatic redo log fetching from the source

Benefits:
• Customers can balance the workload between two different sites while maintaining the Multitenant consolidation
benefits
• The role transition for a single PDB is significantly faster than doing the same at the CDB level

In Oracle Database 23c, Per-PDB Data Guard now supports the standby PDBs being opened in Read-
Only mode to allow offloading of reporting to the DGPDB standby

Copyright © 2023, Oracle and/or its affiliates

Sharding Enhancements
Oracle database 23c provides built-in replication for Oracle Sharding without requiring
configuration of Oracle GoldenGate or Data Guard
How it works:
• Data logically replicated with the consensus-based (Raft) commit protocol enabling APP SERVER

declarative replication configuration, active-active shards, and sub-second failover COORDINATOR

• Makes replication simple and transparent

SHARD 1 SHARD 3

Benefits:
• Meets data sovereignty requirements and supports applications that require low latency and IRELAND
SPAIN
ITALY
GERMANY
FRANCE
USA

high availability
• Supports fast failover with zero data loss, sub-second failover improves availability and SLAs
• Workload balanced among all available hardware as all the shards are symmetric and actively
used, improving hardware utilization

Copyright © 2023, Oracle and/or its affiliates

Upgrade Path to Oracle Database 23c

19
c
g
11

23
c
c
12
c
12
18 c
21 c

Copyright © 2023, Oracle and/or its affiliates

 Oracle Database 23c, Complete & Simple

Copyright © 2023, Oracle and/or its affiliates

Try Everything…for FREE

free-oracle.github.io cloud.oracle.com/free bit.ly/ADB_free developer.oracle.com/

livelabs

Copyright © 2023, Oracle and/or its affiliates

Please rate this session.
Session ID – DB02
Thank you

Copyright © 2023, Oracle and/or its affiliates