Oracle Database 10g Release 1: What's New, Release History & Support Lifecycle

What Is New in Oracle Database 10g Release 1

Oracle Database 10g Release 1 (10.1) marks a fundamental shift in how Oracle positions its database platform -- from a manually administered engine to a self-managing, grid-aware system. The "g" in 10g stands for Grid, and nearly every major feature in this release reflects that design intent: automate the tedious, absorb storage complexity, and give the DBA diagnostic intelligence instead of raw instrumentation.

What is Automatic Storage Management (ASM) in Oracle 10g and how does it replace a volume manager?

Automatic Storage Management (ASM) is a new integrated file system and volume manager introduced in Oracle Database 10g Release 1 that allows the database to manage datafiles, controlfiles, and redo logs directly -- without a third-party logical volume manager or file system product.

In practice, ASM replaces the traditional raw-device or third-party LVM setup with a concept called disk groups. You assign physical disks (or LUNs) to a disk group, and ASM handles striping, mirroring, and rebalancing automatically. When a new disk is added to a disk group, ASM begins rebalancing I/O in the background without any downtime. When a disk fails, ASM automatically re-mirrors the affected extents onto surviving disks.

ASM runs as a separate Oracle instance (with no data dictionary of its own) and introduces three new background processes: ASMB, RBAL, and ARBx. Watch out for the COMPATIBLE.ASM and COMPATIBLE.RDBMS disk group attributes -- they default to 10.1 and must be explicitly advanced to unlock features from later releases. Only RMAN can be used to back up files stored in ASM disk groups.

The syntax to use ASM is straightforward. Reference a disk group with a "+" prefix wherever a filename is expected:

-- Create a tablespace backed by an ASM disk group
CREATE TABLESPACE sales_data
  DATAFILE '+DATA' SIZE 500M AUTOEXTEND ON;

-- Add a disk to an existing disk group (rebalancing is automatic)
ALTER DISKGROUP DATA ADD DISK '/dev/sdd1';

-- Check rebalance progress
SELECT GROUP_NUMBER, OPERATION, STATE, POWER, SOFAR, EST_WORK
FROM   V$ASM_OPERATION;

Most teams migrating from 9i to 10g adopt ASM incrementally -- keeping existing cooked file systems and moving only new databases or new tablespaces to ASM. The migration path from a disk-based backup to full ASM storage is well-documented and does not require a full cold migration.

How does Oracle 10g automate performance diagnostics with AWR, ADDM, and ASH?

Oracle Database 10g Release 1 introduces a trio of tightly integrated self-tuning components -- the Automatic Workload Repository (AWR), the Automatic Database Diagnostic Monitor (ADDM), and Active Session History (ASH) -- that together replace manual Statspack workflows and ad-hoc tuning sessions.

The Automatic Workload Repository (AWR) is the successor to Statspack. It automatically captures, processes, and retains performance statistics at configurable intervals (default: every 60 minutes, retained for 7 days). Unlike Statspack, AWR is integrated into the kernel and feeds all other advisory components. The underlying data is stored in the new mandatory SYSAUX tablespace.

The Automatic Database Diagnostic Monitor (ADDM) runs automatically after each AWR snapshot and analyzes the collected data to identify the most significant performance bottleneck in the previous interval. Rather than presenting raw wait statistics, ADDM produces ranked findings with specific, actionable recommendations -- whether that is adding an index, enabling parallel query, or increasing the buffer cache. This matters if your team previously spent hours correlating Statspack reports by hand.

The Active Session History (ASH) samples active sessions every second and retains the in-memory history for real-time analysis. It fills the gap between AWR snapshots, enabling sub-minute diagnosis of transient problems that would otherwise vanish before the next AWR report.

-- Run an ADDM report for a specific AWR snapshot range
DECLARE
  task_name  VARCHAR2(30) := 'my_addm_task';
  begin_snap NUMBER       := 100;
  end_snap   NUMBER       := 101;
BEGIN
  DBMS_ADVISOR.CREATE_TASK(
    advisor_name => 'ADDM',
    task_name    => task_name
  );
  DBMS_ADVISOR.SET_TASK_PARAMETER(task_name, 'START_SNAPSHOT', begin_snap);
  DBMS_ADVISOR.SET_TASK_PARAMETER(task_name, 'END_SNAPSHOT',   end_snap);
  DBMS_ADVISOR.EXECUTE_TASK(task_name);
END;
/

-- Or use the supplied script directly (simpler in most cases)
@$ORACLE_HOME/rdbms/admin/addmrpt.sql

In addition to ADDM, the SQL Tuning Advisor works with the Automatic Tuning Optimizer (ATO) to analyze high-load SQL from the AWR workload history and generate recommendations for new indexes, SQL profiles, or query restructuring -- all without manual intervention unless a recommendation requires a schema change.

Automatic SGA Management (ASMM) completes the picture by dynamically resizing the shared pool, buffer cache, and other SGA components based on real-time workload data from AWR, eliminating the need to manually tune DB_CACHE_SIZE and SHARED_POOL_SIZE for most workloads.

What flashback features does Oracle 10g Release 1 introduce for accidental data recovery?

Oracle Database 10g Release 1 dramatically expands Flashback Technology beyond the row-level Flashback Query introduced in 9i, adding Flashback Database, Flashback Table, Flashback Drop, and Flashback Version Query -- each targeting a different tier of accidental loss.

Flashback Database allows you to rewind the entire database to a past point in time using flashback logs stored in the Flash Recovery Area, without restoring from backup. This is orders of magnitude faster than a conventional point-in-time recovery (PITR) via RMAN for most scenarios. Watch out: flashback logs must be enabled before the incident, and the Flash Recovery Area must be appropriately sized to retain enough history.

Flashback Drop introduces an Oracle Recycle Bin. When a table is dropped without the PURGE keyword, 10g renames the segment and places it in the recycle bin rather than immediately deallocating space. You can query the recycle bin and restore the table -- with its indexes and constraints -- using a single command. Space is reclaimed only when the database needs it.

-- Show what is in the recycle bin for the current user
SHOW RECYCLEBIN;

-- Restore a dropped table (optionally rename it on recovery)
FLASHBACK TABLE orders TO BEFORE DROP RENAME TO orders_recovered;

-- Permanently drop, bypassing the recycle bin
DROP TABLE orders PURGE;

-- Flash the entire database back to a restore point (requires MOUNT mode)
STARTUP MOUNT;
FLASHBACK DATABASE TO RESTORE POINT before_batch_load;
ALTER DATABASE OPEN RESETLOGS;

Flashback Table rewinds a live, online table to a past SCN or timestamp using undo data -- no downtime, no export/import cycle. Row Movement must be enabled on the table first. Flashback Version Query uses the VERSIONS BETWEEN clause to show every version a specific row passed through during a time range, including the transaction ID responsible for each change -- invaluable for root-cause analysis on data corruption issues.

In practice, these features reduce the blast radius of runaway batch jobs, errant DML, and developer mistakes in production environments. Many teams retire their "export before the batch" safety scripts once Flashback Database and Flashback Drop are properly configured.

What replaces the original EXP and IMP utilities in Oracle 10g and why does it matter?

Oracle Data Pump -- invoked via the expdp and impdp command-line clients -- replaces the original exp and imp utilities in Oracle Database 10g Release 1, delivering significantly faster bulk data movement through a server-side architecture and direct path I/O.

The key architectural difference is that Data Pump operations execute inside the database server, not in the client process. This eliminates the client-server network bottleneck that plagued large exp/imp jobs and makes parallel execution practical. A PARALLEL parameter controls the degree of parallelism directly.

-- Export a schema with 4 parallel workers, compressed
expdp hr/hr SCHEMAS=hr DIRECTORY=DATA_PUMP_DIR \
      DUMPFILE=hr_%U.dmp PARALLEL=4 LOGFILE=hr_exp.log

-- Import into a different schema, remapping
impdp system/manager DIRECTORY=DATA_PUMP_DIR \
      DUMPFILE=hr_%U.dmp REMAP_SCHEMA=hr:hr_test \
      LOGFILE=hr_imp.log

-- Estimate export size without writing data
expdp hr/hr SCHEMAS=hr ESTIMATE_ONLY=YES

Data Pump also introduces fine-grained filtering via INCLUDE and EXCLUDE parameters, the ability to remap tablespaces and schemas at import time, and an interactive attach mode that lets you monitor or modify a running job without cancelling it. Cross-Platform Transportable Tablespaces are also extended in 10g to support movement between different operating system platforms -- a major improvement over 9i which restricted tablespace transport to the same platform.

This matters if your team runs nightly logical backups via exp or performs regular schema refreshes between environments. The original exp/imp utilities remain available in 10g for backward compatibility but should be considered legacy; they are not enhanced in this or subsequent releases.

What SQL and PL/SQL developer features are new in Oracle Database 10g Release 1?

Oracle Database 10g Release 1 adds a substantial set of SQL and PL/SQL enhancements that directly reduce application-level complexity, with Regular Expression support and DBMS_SCHEDULER being the two changes most likely to affect day-to-day development.

Regular Expressions in SQL and PL/SQL arrive via four new functions: REGEXP_LIKE, REGEXP_INSTR, REGEXP_SUBSTR, and REGEXP_REPLACE. These use POSIX Extended Regular Expression (ERE) syntax and integrate with NLS settings through the NLS_COMP=LINGUISTIC parameter, which ensures consistent linguistic comparison behaviour across all SQL string operations.

-- Validate email format using REGEXP_LIKE
SELECT email
FROM   customers
WHERE  REGEXP_LIKE(email, '^[A-Za-z0-9._%+-]+@[A-Za-z0-9.-]+\.[A-Za-z]{2,}$');

-- Replace multiple consecutive spaces with a single space
SELECT REGEXP_REPLACE('too   many   spaces', ' {2,}', ' ') AS cleaned
FROM   DUAL;

-- Extract the domain part of a URL
SELECT REGEXP_SUBSTR('https://www.example.com/page', 'www\.[^/]+') AS domain
FROM   DUAL;

DBMS_SCHEDULER replaces the aging DBMS_JOB package with a production-grade job scheduling framework. Unlike DBMS_JOB, DBMS_SCHEDULER supports named programs and schedules, job classes with resource manager integration, OS command execution (shell scripts, executables), calendar-based repeat intervals, and event-driven execution. Most teams should migrate existing DBMS_JOB logic to DBMS_SCHEDULER during any 9i-to-10g upgrade.

Additional developer-facing additions include: UTL_MAIL for simple SMTP email from PL/SQL, UTL_COMPRESS for gzip-compatible compression of RAW and BLOB data, the PL/SQL optimizing compiler (which reorders and reorganizes code at compile time for better runtime performance), and compiler warning messages surfaced via DBMS_WARNING. The MERGE statement is also enhanced to support DELETE in the WHEN MATCHED clause and to allow unconditional INSERT without a matching source row.