Unlocking Oracle 23c AI's JSON Relational Duality: Simplifying Data Handling for Modern Applications

Unlocking Oracle 23c AI's JSON Relational Duality: Simplifying Data Handling for Modern Applications

With the release of Oracle 23c AI, Oracle has introduced a suite of new features designed to empower DBAs and developers. One of the most revolutionary additions is JSON Relational Duality, which offers unprecedented flexibility in handling data by allowing users to access the same data as both relational and JSON, seamlessly switching between formats without duplication or additional processing.

This article explores JSON Relational Duality and demonstrates how it simplifies data management for applications that need to handle both structured and unstructured data. 

What Is JSON Relational Duality?

JSON Relational Duality in Oracle 23c AI allows the same data to be represented as both relational and JSON simultaneously. This feature is particularly valuable for applications where structured relational data needs to be accessed as unstructured JSON data—and vice versa. 

For example, a company with a traditional relational structure may want to integrate with a modern application that uses JSON-based APIs. JSON Relational Duality bridges this gap, enabling applications to leverage both formats without complex transformations or performance compromises.

Why JSON Relational Duality Matters

In the past, handling both relational and JSON data often meant duplicating data or creating intricate ETL processes to convert between formats. JSON Relational Duality eliminates these challenges by allowing a single, consistent view of data that works natively with both JSON and relational formats.

- Reduced Complexity: You avoid complex ETL processes when transforming data formats.

- Enhanced Flexibility: Developers can work with the format best suited to their application needs.

- Improved Performance: Access data in the format you need without extra processing, reducing load on the system.

Getting Started with JSON Relational Duality

To demonstrate JSON Relational Duality, let’s use an example of a user, 'user', who manages employee data in a relational structure but needs to provide a JSON format for API consumption.

Suppose we have a table 'employees' with the following schema:

SQL

CREATE TABLE employees (

    employee_id NUMBER PRIMARY KEY,

    first_name VARCHAR2(50),

    last_name VARCHAR2(50),

    department VARCHAR2(50),

    salary NUMBER

);

Step 1: Enabling JSON Relational Duality

With Oracle 23c AI, we can use a 'DUPLICATE JSON' clause to enable duality, allowing 'employees' to be queried as JSON without changing the schema.

SQL

ALTER TABLE employees ENABLE JSON RELATIONAL DUALITY;

Step 2: Querying Data in JSON Format

Now, 'user' can retrieve the same data in JSON format using a simple query:

SQL

SELECT JSON_OBJECT(*) 

FROM employees 

WHERE department = 'Sales';

The query above will output each row as a JSON object, making it compatible with any API or application that requires JSON.

Step 3: Using JSON Data with Relational Functions

JSON Relational Duality allows JSON data to be used in SQL operations as if it were relational. This means that the 'user' can join, filter, and manipulate JSON data using SQL.

For example, let’s filter employees by salary and project it as JSON:

SQL

SELECT JSON_OBJECT(*) 

FROM employees 

WHERE salary > 70000;

Use Case: Integrating Relational Data with a JSON-Based Application

Imagine the 'user' is tasked with integrating an Oracle database with a cloud-based application that consumes JSON via REST API. JSON Relational Duality simplifies this process. Instead of transforming data into JSON on the fly or storing duplicates, the 'user' can directly query relational data in JSON format and feed it into the API, streamlining integration.

Example API Call with JSON Duality Data

Using JSON output from Oracle, the 'user' can now make API calls without extra data transformation:

JSON:

{

    "employee_id": 101,

    "first_name": "John",

    "last_name": "Doe",

    "department": "Sales",

    "salary": 75000

}

 

JSON Relational Duality in Complex Queries

Let’s explore a more complex query in which the 'user' needs to get department-wise average salaries, outputting the results in JSON format.

SQL

SELECT JSON_OBJECT(

    'department' VALUE department,

    'average_salary' VALUE AVG(salary)

) AS department_summary

FROM employees

GROUP BY department;

This query enables a modern, JSON-based interface to summarize data traditionally stored in relational format, making it accessible to front-end applications that prefer JSON.

SEO-Optimized Benefits of JSON Relational Duality

For DBAs, developers, and system architects, JSON Relational Duality in Oracle 23c AI is a keyword-rich feature that aligns with trends in hybrid database management and Oracle 23c AI JSON integration. It emphasizes dual-format data handling and native JSON support in relational databases, making it ideal for applications with mixed data needs. Key benefits like performance optimization, reduced data duplication, and simplified data management address frequently searched queries, ensuring that Oracle 23c AI remains at the forefront of modern database management.

Summary: Embracing the Duality Advantage in Oracle 23c AI

Oracle 23c AI’s JSON Relational Duality is a groundbreaking feature, particularly for DBAs like 'user', who must manage complex data structures efficiently. With JSON Relational Duality, Oracle offers flexibility without sacrificing performance, enabling seamless data integration for applications that demand both structured and unstructured data.

For anyone managing databases in hybrid environments, JSON Relational Duality is a powerful tool that reduces workload, enhances performance, and provides a competitive edge in the data-driven landscape.

Oracle EXPDP commands

The expdp (Data Pump Export) utility is used to export data and metadata from an Oracle database. It provides several command-line options to customize the export process. Below are some common expdp commands and their explanations:

Basic Syntax:
 

expdp [username/password] DIRECTORY=directory_name DUMPFILE=dumpfile_name LOGFILE=logfile_name [options]

Common Expdp Commands:
Export a Full Database:
expdp system/password FULL=Y DIRECTORY=dpump_dir DUMPFILE=full_db.dmp LOGFILE=full_db.log

FULL=Y: Exports the entire database.
Export a Specific Schema:
expdp system/password SCHEMAS=schema_name DIRECTORY=dpump_dir DUMPFILE=schema.dmp LOGFILE=schema.log

SCHEMAS=schema_name: Exports a specific schema.
Export Specific Tables:
expdp system/password TABLES=table1,table2 DIRECTORY=dpump_dir DUMPFILE=tables.dmp LOGFILE=tables.log

TABLES=table1,table2: Exports specific tables.
Export a Specific Table with Data and Metadata:

expdp system/password TABLES=table_name DIRECTORY=dpump_dir DUMPFILE=table.dmp LOGFILE=table.log

TABLES=table_name: Exports a specific table.
Export with Compression:
expdp system/password FULL=Y DIRECTORY=dpump_dir DUMPFILE=full_db.dmp LOGFILE=full_db.log COMPRESSION=ALL

COMPRESSION=ALL: Compresses all data during export.
Export with Data Filtering (e.g., Export Data from a Specific Date):
expdp system/password TABLES=table_name DIRECTORY=dpump_dir DUMPFILE=table.dmp LOGFILE=table.log QUERY=table_name:"WHERE created_date > TO_DATE('2024-01-01', 'YYYY-MM-DD')"
QUERY=table_name:"WHERE condition": Filters rows based on a condition.
Export Metadata Only:
expdp system/password FULL=Y DIRECTORY=dpump_dir DUMPFILE=metadata.dmp LOGFILE=metadata.log CONTENT=METADATA_ONLY
CONTENT=METADATA_ONLY: Exports only metadata (no data).
Export Data Only:
expdp system/password FULL=Y DIRECTORY=dpump_dir DUMPFILE=data_only.dmp LOGFILE=data_only.log CONTENT=DATA_ONLY
CONTENT=DATA_ONLY: Exports only data (no metadata).
Export a Database with a Specific Date Format:
expdp system/password FULL=Y DIRECTORY=dpump_dir DUMPFILE=full_db.dmp LOGFILE=full_db.log PARALLEL=4
PARALLEL=4: Uses 4 parallel threads for faster export.
Export with a Job Name:

expdp system/password FULL=Y DIRECTORY=dpump_dir DUMPFILE=full_db.dmp LOGFILE=full_db.log JOB_NAME=export_full_db
JOB_NAME=export_full_db: Assigns a name to the export job.

Additional Parameters:
CONTENT: Specifies whether to export metadata only (METADATA_ONLY), data only (DATA_ONLY), or both (ALL).
EXCLUDE: Excludes specific objects or object types from the export. Example: EXCLUDE=TABLE:"='table_name'".
INCLUDE: Includes specific objects or object types in the export. Example: INCLUDE=TABLE:"IN ('table1', 'table2')".
REMAP_SCHEMA: Remaps schema names. Example: REMAP_SCHEMA=old_schema:new_schema.
REMAP_TABLESPACE: Remaps tablespace names. Example: REMAP_TABLESPACE=old_tablespace:new_tablespace.

Directory Object:
Before running expdp, ensure that the DIRECTORY object exists in the database and points to a valid filesystem directory where the dump files will be written.
 
 
CREATE OR REPLACE DIRECTORY dpump_dir AS '/path/to/directory';

Example Execution:
To execute an expdp command, open a command prompt or terminal and run the appropriate expdp command based on your requirements. Ensure you have the necessary privileges and that the Oracle environment variables (ORACLE_HOME and PATH) are set correctly.


Conclusion:
The expdp utility offers powerful options for exporting data and metadata from Oracle databases. By using the appropriate parameters and options, you can tailor the export process to meet specific needs and optimize performance.

Identify and Terminate Sessions in oracle

 

First, you have to identify the session to be killed with alter system kill session.

Step-by-Step Instructions to Identify and Terminate a Session:

Invoke SQL*Plus:

First, open SQL*Plus to begin the process.

Query V$SESSION to Identify the Session:

Use the following query to retrieve the session details. This will list all active sessions with their respective SID, SERIAL#, STATUS, SCHEMANAME, and PROGRAM.

SELECT SID, SERIAL#, STATUS, SCHEMANAME, PROGRAM FROM v$session;

The SID (Session Identifier) and SERIAL# values of the Oracle session to be killed can then be identified from this output.

Execute the ALTER SYSTEM Command:

Substitute the identified SID and SERIAL# values and issue the alter system kill session command.

ALTER SYSTEM KILL SESSION 'sid,serial#';

Handling Sessions Marked for Kill:

Sometimes, Oracle is not able to kill the session immediately with the alter system kill session command alone. Upon issuing the command, the session may be 'marked for kill' and will be terminated as soon as possible.

Forcing Session Termination:

In the case where a session is 'marked for kill' and not terminated immediately, you can force the termination by adding the immediate keyword:

ALTER SYSTEM KILL SESSION 'sid,serial#' IMMEDIATE;

Verify Termination:

To ensure that the session has been terminated, re-query the V$SESSION dynamic performance view:

SELECT SID, SERIAL#, STATUS, SCHEMANAME, PROGRAM FROM v$session WHERE SID = 'sid' AND SERIAL# = 'serial#';

The value of the STATUS column will be 'ACTIVE' when the session is making a SQL call and 'INACTIVE' if it is not.

Confirm PMON Cleanup:

After the Process Monitor (PMON) has cleaned up the session, the row will be removed from V$SESSION. Re-query to confirm:

SELECT SID, SERIAL#, STATUS, SCHEMANAME, PROGRAM FROM v$session WHERE SID = 'sid' AND SERIAL# = 'serial#';

Handling RAC Environments:

In a Real Application Clusters (RAC) environment, you can optionally specify the INST_ID, which is shown when querying the GV$SESSION view. This allows you to kill a session on a different RAC node.

ALTER SYSTEM KILL SESSION 'sid,serial#,@inst_id';

By following these detailed steps, you can efficiently identify and terminate a session using SQL*Plus, ensuring minimal disruption and maintaining database integrity.

Oracle Resource Manager: Granular Control for Managing Idle Sessions

Oracle Resource Manager: Granular Control for Managing Idle Sessions

In database management, efficiently handling idle sessions is essential to maintaining performance and resource availability. Oracle Resource Manager is a powerful tool, often underutilized, that offers granular control over how sessions consume resources. This post explores how to leverage Oracle Resource Manager to monitor and automatically terminate idle sessions, complete with in-depth, practical examples.

Why Use Oracle Resource Manager?

Oracle Resource Manager allows DBAs to define and enforce resource limits on user sessions based on specific conditions, such as idle time, CPU usage, and session priority. This level of control can be particularly useful in environments with high session volumes, such as transactional systems or shared database infrastructures, where idle sessions can prevent other active sessions from connecting.

Setting Up Resource Manager for Idle Session Management

To illustrate, let's walk through a scenario where we have a transactional database with frequent user connections. Our goal is to manage idle sessions, terminating any session that remains idle for over an hour. 

Step 1: Creating a Resource Plan

A resource plan acts as a blueprint, defining how Oracle should handle session resources. We’ll create a resource plan named `Session_Management_Plan` to automatically switch idle sessions to a termination state.

SQL:

BEGIN

    DBMS_RESOURCE_MANAGER.create_pending_area();

    DBMS_RESOURCE_MANAGER.create_plan(

        plan    => 'Session_Management_Plan',

        comment => 'Resource Plan to handle idle sessions exceeding 1 hour'

    );

    DBMS_RESOURCE_MANAGER.submit_pending_area();

END;

/

Here, we set up a simple resource plan structure to act as a container for directives. The `create_pending_area` function places the configuration in a pending state until it is finalized and submitted.

Step 2: Defining Consumer Groups

In Oracle Resource Manager, consumer groups classify sessions based on their resource needs. Here, we create two consumer groups: `ACTIVE_SESSIONS` for sessions with standard activity and `IDLE_TERMINATION` for sessions that exceed the idle limit.

SQL:

BEGIN

    DBMS_RESOURCE_MANAGER.create_pending_area();

    DBMS_RESOURCE_MANAGER.create_consumer_group(

        consumer_group => 'ACTIVE_SESSIONS',

        comment        => 'Group for actively monitored sessions'

    );

    DBMS_RESOURCE_MANAGER.create_consumer_group(

        consumer_group => 'IDLE_TERMINATION',

        comment        => 'Group for sessions to be terminated if idle too long'

    );

    DBMS_RESOURCE_MANAGER.submit_pending_area();

END;

/

These consumer groups allow us to specify different rules and behaviors based on session status.

Step 3: Creating Plan Directives

Directives define the rules for each consumer group within a plan. We’ll set a directive to monitor sessions within `ACTIVE_SESSIONS`, moving idle sessions to the `IDLE_TERMINATION` group if they remain idle for over an hour.

SQL:

BEGIN

    DBMS_RESOURCE_MANAGER.create_pending_area();

    DBMS_RESOURCE_MANAGER.create_plan_directive(

        plan             => 'Session_Management_Plan',

        group_or_subplan => 'ACTIVE_SESSIONS',

        comment          => 'Standard sessions with monitoring for idle state',

        switch_time      => 3600, -- 1 hour idle limit (3600 seconds)

        switch_group     => 'IDLE_TERMINATION'

    );

    DBMS_RESOURCE_MANAGER.create_plan_directive(

        plan             => 'Session_Management_Plan',

        group_or_subplan => 'IDLE_TERMINATION',

        comment          => 'Group for terminating sessions idle over 1 hour',

        cpu_p1           => 0, -- No CPU allocation as sessions are idle

        max_idle_blocker_time => 3600 -- Terminate after one hour idle

    );

    DBMS_RESOURCE_MANAGER.submit_pending_area();

END;

/

In this configuration:

- `switch_time` specifies the idle duration threshold (3600 seconds or 1 hour).

- `switch_group` moves sessions from `ACTIVE_SESSIONS` to `IDLE_TERMINATION` once they exceed the idle time.

- `cpu_p1` is set to zero for `IDLE_TERMINATION` to prevent idle sessions from consuming CPU.

- `max_idle_blocker_time` limits the maximum time for idle sessions in the `IDLE_TERMINATION` group, ensuring termination.

Step 4: Activating the Resource Plan

With the resource plan and directives set, we activate `Session_Management_Plan` to enforce these rules on the database.

SQL:

ALTER SYSTEM SET resource_manager_plan = 'Session_Management_Plan';

 

By activating this plan, Oracle will apply our specified rules for monitoring idle sessions and automatically terminate them after an hour of inactivity.

Step 5: Verifying and Monitoring Session Status

To ensure the plan is working as expected, monitor session activity by checking which sessions are in `ACTIVE_SESSIONS` versus `IDLE_TERMINATION`.

SQL:

SELECT username, 

       session_id, 

       status, 

       last_call_et AS idle_seconds, 

       consumer_group

FROM   v$session

WHERE  consumer_group IN ('ACTIVE_SESSIONS', 'IDLE_TERMINATION')

ORDER BY idle_seconds DESC;

The `last_call_et` column shows idle time in seconds, while the `consumer_group` column lets you see whether the session is active or set for termination. This query provides visibility into session status, letting you track how effectively idle sessions are managed.

Additional Tips for Using Oracle Resource Manager

1. Testing in Non-Production: Resource Manager settings should be tested in non-production environments first to ensure they align with your workload and that CPU or memory-intensive processes are unaffected by idle session policies.

2. Avoid Overly Aggressive Termination Policies: While terminating idle sessions frees up resources, be cautious with overly aggressive thresholds that could disrupt user sessions critical for business operations.

3. Regularly Review Session Activity: Use `DBA_HIST_ACTIVE_SESS_HISTORY` or `DBA_HIST_RESOURCE_PLAN_DIRECTIVE` views to analyze historical session activity and refine thresholds as needed.

Conclusion:

Oracle Resource Manager offers a structured, policy-based way to manage session resources, making it easier for DBAs to maintain optimal database performance. By automatically handling idle sessions, DBAs can prevent resource bottlenecks and maximize availability for active sessions. With careful setup and monitoring, Resource Manager becomes an invaluable tool in any Oracle DBA’s toolkit.

References: Oracle Database Administrator’s Guide, Oracle 19c Documentation, Oracle Database Resource Manager Concepts and Usage.

Managing and Terminating Idle Oracle Database Sessions Automatically

 

Managing and Terminating Idle Oracle Database Sessions Automatically

In database environments, maintaining optimal performance requires careful management of user sessions, especially idle ones. Idle sessions can consume resources and lead to unnecessary wait times, especially when the maximum number of sessions is reached. In this post, we’ll explore how to identify idle sessions, create a configuration to manage them, and set up an automatic process to terminate idle sessions that last longer than one hour.

Understanding Idle Sessions in Oracle Database

An idle session is a session where the user is connected but not actively executing SQL statements. While Oracle can handle many sessions simultaneously, excessive idle sessions can eventually strain resources. By managing and terminating these sessions, DBAs can optimize resource usage, particularly in environments where there’s a limit on concurrent connections.

Method 1: Using Oracle Profiles to Manage Idle Sessions

Oracle Profiles allows DBAs to control session-level resource usage, including limiting the amount of idle time.

Step 1: Create a Profile with an Idle Time Limit

To set an idle timeout of one hour (60 minutes), we’ll create a profile with a `IDLE_TIME` resource limit. 

SQL: 

CREATE PROFILE user_idle_limit_profile

LIMIT

    IDLE_TIME 60; -- 60 minutes

Here, `IDLE_TIME` represents the maximum allowed idle time for a session, in minutes. Once this threshold is exceeded, Oracle will automatically disconnect the session.

Step 2: Assign the Profile to Users

To enforce the idle timeout, assign the profile to users who require monitoring:

SQL: 

ALTER USER john_doe PROFILE user_idle_limit_profile;

ALTER USER jane_smith PROFILE user_idle_limit_profile;

All users assigned to `user_idle_limit_profile` will be disconnected automatically if idle for more than one hour. Keep in mind that an `IDLE_TIME` of zero means there is no idle limit.

 Note: Sessions terminated this way will receive an ORA-02396 warning message, indicating that they have been disconnected due to inactivity.

Step 3: Verifying Profile Application

You can verify if the profile was applied correctly by checking the `DBA_USERS` view:

SQL: 

SELECT username, profile 

FROM dba_users 

WHERE profile = 'USER_IDLE_LIMIT_PROFILE';

This confirms that the intended users are bound by the idle session limit.

Method 2: Using Oracle Resource Manager for More Granular Control

Oracle’s Resource Manager offers more control. It allows you to create rules that monitor and act on specific session conditions, including idle sessions.

Step 1: Create a Resource Plan

First, create a resource plan that specifies the action to take when a session remains idle for too long. Here’s an example:

SQL: 

BEGIN

    DBMS_RESOURCE_MANAGER.create_pending_area();

    DBMS_RESOURCE_MANAGER.create_plan(

        plan    => 'idle_session_plan',

        comment => 'Resource plan to terminate idle sessions'

    );

    DBMS_RESOURCE_MANAGER.create_plan_directive(

        plan             => 'idle_session_plan',

        group_or_subplan => 'OTHER_GROUPS',

        comment          => 'Terminate sessions idle for more than 60 minutes',

        switch_for_call  => 'TRUE',

        switch_time      => 3600, -- 3600 seconds (1 hour)

        switch_group     => 'KILL_SESSION'

    );

    DBMS_RESOURCE_MANAGER.submit_pending_area();

END;

/

In this example, the `switch_time` parameter is set to 3600 seconds (1 hour). If a session remains idle for this duration, Oracle will automatically move it to the `KILL_SESSION` group, where it’s terminated.

Step 2: Activate the Resource Plan

Once the resource plan is defined, activate it to enforce the rules:

SQL: 

ALTER SYSTEM SET resource_manager_plan = 'idle_session_plan';

With this plan active, sessions that remain idle for over an hour are automatically terminated, reducing unnecessary resource consumption.

Step 3: Verifying Active Resource Plan

To check which resource plan is active, you can use:

SQL: 

SELECT name, active

FROM v$rsrc_plan

WHERE is_top_plan = 'TRUE';

This will confirm that `idle_session_plan` is the active resource management plan.

Monitoring and Auditing Idle Sessions

Regularly reviewing the sessions terminated by the profile or resource manager helps ensure that active users are unaffected. You can monitor terminated sessions by querying Oracle’s session and resource views:

SQL: 

SELECT sid, serial#, username, status, last_call_et

FROM v$session

WHERE status = 'INACTIVE' 

  AND last_call_et > 3600; -- Sessions idle for more than 1 hour

The `last_call_et` column records the idle duration in seconds. This query allows you to keep track of all sessions that have been idle for over an hour.

Final Thoughts and Best Practices

1. Set Realistic Limits: Configure idle timeouts that reflect actual user behavior. For instance, if users typically require 30-minute breaks, a 60-minute idle timeout may be ideal.

2. Monitor Frequently: As with any automated process, regularly monitor your configuration to ensure that critical sessions are not mistakenly terminated.

3. Communicate with Users: Inform users about session policies, especially if implementing a more aggressive timeout.

By setting up automated processes to manage idle sessions, Oracle DBAs can improve resource utilization, ensure a smooth user experience, and optimize system performance. This proactive approach helps avoid potential bottlenecks, allowing DBAs to focus on higher-impact tasks.

Optimizing Storage and Performance with Oracle Database Advanced Compression

 Optimizing Storage and Performance with Oracle Database Advanced Compression

In the world of database management, the challenge of balancing performance with storage efficiency is ever-present. Oracle's Advanced Compression feature offers a robust solution to this by reducing storage footprint, optimizing performance, and saving costs — all without compromising data integrity. In this post, we’ll dive into how you can leverage advanced compression in Oracle, with practical examples and straightforward code snippets.

  Why Use Advanced Compression in Oracle?

Advanced Compression isn't just about saving storage; it enhances data retrieval speed by minimizing the amount of data that needs to be read from disk. This is especially beneficial for larger tables or indices that experience heavy I/O operations. Advanced Compression works across various Oracle database objects and data types, from tables and indexes to LOBs and backups.

  Types of Compression in Oracle Database

Oracle provides several types of compression suited to various workloads:

1. Basic Table Compression: Suitable for read-only or static tables, it compresses data as it's loaded.

2. OLTP Table Compression: Designed for transactional environments, it works dynamically on data inserts and updates.

3. Hybrid Columnar Compression (HCC): Available only on Exadata and a few other Oracle storage solutions, it’s ideal for data warehouses and archival tables.

4. Index Compression: Reduces the storage for indexes by compressing repeated values.

Let’s walk through these options with examples using a hypothetical table and index.

  1. Basic Table Compression

To start, let’s look at Basic Table Compression, commonly used for data warehousing or infrequently updated tables.

Suppose we have a table `Employee_Records` that stores historical data and is rarely modified:

 SQL :

CREATE TABLE Employee_Records (

    employee_id NUMBER,

    first_name VARCHAR2(50),

    last_name VARCHAR2(50),

    department_id NUMBER,

    hire_date DATE

) COMPRESS BASIC;

 

Here, we specify `COMPRESS BASIC`, which instructs Oracle to compress the data during bulk insert operations. Keep in mind that this is ideal for tables where data modifications are infrequent, as Basic Compression doesn’t automatically recompress on updates.

  2. OLTP Table Compression

For transactional tables with frequent updates, OLTP Table Compression is more appropriate. Oracle's advanced algorithm ensures that as rows are inserted or updated, the data remains compressed without additional manual intervention.

Let’s use `Order_Details`, a table holding real-time sales data:

 SQL :

CREATE TABLE Order_Details (

    order_id NUMBER,

    product_id NUMBER,

    quantity NUMBER,

    order_date DATE,

    customer_id NUMBER

) COMPRESS FOR OLTP;

 

With `COMPRESS FOR OLTP`, this table will dynamically compress data during both inserts and updates, keeping storage usage low while maintaining fast performance.

Tip: OLTP compression is a powerful way to balance performance and space in high-transaction environments. However, it requires some CPU overhead, so monitor performance in CPU-bound systems.

  3. Hybrid Columnar Compression (HCC)

HCC is an exclusive feature for Exadata, ZFS, and Oracle Cloud customers. It combines row and column storage, making it highly efficient for queries that scan large datasets. HCC can offer multiple levels, like `COMPRESS FOR QUERY` and `COMPRESS FOR ARCHIVE`, allowing you to optimize for retrieval speed or maximum storage savings.

Using `Sales_Archive` as an example table:

 SQL :

CREATE TABLE Sales_Archive (

    sale_id NUMBER,

    region VARCHAR2(30),

    sale_amount NUMBER,

    sale_date DATE

) COMPRESS FOR QUERY HIGH;

 

In this case, `COMPRESS FOR QUERY HIGH` balances both storage efficiency and query performance, ideal for analytic queries on historical data. For long-term storage where frequent access isn’t necessary, consider `COMPRESS FOR ARCHIVE HIGH`.

  4. Index Compression

Finally, Index Compression can be particularly beneficial for large indexes with repeating values. By compressing the key prefix, Oracle significantly reduces index size, speeding up queries.

Imagine we have a commonly queried index on the `Employee_Records` table:

 SQL :

CREATE INDEX idx_employee_department 

ON Employee_Records (department_id, last_name, first_name) COMPRESS 1;

 

In this example, `COMPRESS 1` compresses the leading column (`department_id`) in the index. If more columns contain duplicate values, you can increase the compression level. Index compression works well in cases where there’s repetition, such as departmental groupings, making it a great fit for HR or CRM databases.

  Monitoring and Verifying Compression Benefits

To see the benefits of compression, you can query Oracle's data dictionary to monitor the space savings:

 SQL :

SELECT segment_name, 

       segment_type, 

       bytes / 1024 / 1024 AS size_MB 

FROM   dba_segments 

WHERE  segment_name = 'EMPLOYEE_RECORDS';

 

By running this query before and after enabling compression, you can quantify the storage savings directly.

  Considerations and Best Practices

1. Testing: Test compression in a non-production environment to assess CPU overhead and performance impact.

2. Monitoring: Regularly monitor space and CPU metrics to ensure that compression is meeting your goals without overtaxing system resources.

3. Reorganizing Tables: Consider periodically reorganizing heavily updated tables with OLTP compression to maintain optimal storage savings.

  Conclusion

Oracle's Advanced Compression provides a flexible toolkit for database administrators looking to optimize both storage and performance. By selecting the right type of compression for each workload, you can achieve substantial space savings while maintaining fast access times. Try implementing compression gradually, starting with non-critical tables, and measure the impact to find the best fit for your database environment.