Oracle Service and Log Operations

Purpose: Guide to managing Oracle database services, listeners, and log files with OraDBA.

Audience: DBAs and system administrators managing Oracle services and log files.

Service Management Overview

OraDBA provides comprehensive service management tools for Oracle databases and listeners, suitable for interactive use, automation, and integration with systemd or init.d. The Plugin System enables product-specific service management for databases, Data Safe connectors, and other Oracle products.

The service management toolkit consists of three main scripts and a root wrapper:

• oradba_dbctl.sh: Database instance control (start/stop/restart/status)
• oradba_lsnrctl.sh: Listener control (start/stop/restart/status)
• oradba_services.sh: Orchestrates both databases and listeners
• oradba_services_root.sh: Root wrapper for systemd/init.d integration

Convenient Aliases

OraDBA provides short aliases for daily use:

# Database control
dbctl           # Alias for oradba_dbctl.sh
dbstart         # Start databases
dbstop          # Stop databases
dbrestart       # Restart databases

# Listener control
listener        # Alias for oradba_lsnrctl.sh
lsnrstart       # Start listeners (oradba_lsnrctl.sh wrapper)
lsnrstop        # Stop listeners (oradba_lsnrctl.sh wrapper)
lsnrstatus      # Listener status (oradba_lsnrctl.sh wrapper)

# Combined services
orastart        # Start all Oracle services
orastop         # Stop all Oracle services
orarestart      # Restart all Oracle services
orastatus       # Show all service status

Database Control

oradba_dbctl.sh controls Oracle database instances with environment-aware execution.

Features

• Honors :Y flag in oratab for auto-start databases
• Explicit SID override supported (ignores :N flag when SID specified)
• Configurable shutdown timeout (default: 180s)
• Escalates from SHUTDOWN IMMEDIATE to SHUTDOWN ABORT on timeout
• Requires justification when stopping ALL databases
• Optional explicit PDB opening
• Continues on errors (logs failures, processes remaining)
• Integrates with oraenv.sh for per-database environment

Usage and Options

oradba_dbctl.sh {start|stop|restart|status} [OPTIONS] [SID1 SID2 ...]

Option	Long Form	Description
-f	--force	Skip confirmation prompts
-t	--timeout SECONDS	Shutdown timeout (default: 180)
-p	--open-pdbs	Explicitly open all PDBs after startup
-h	--help	Show help message

Key Examples

# Start all databases with :Y flag in oratab
oradba_dbctl.sh start

# Start specific databases (overrides :N flag)
oradba_dbctl.sh start ORCL CDB1

# Stop all with 300s timeout
oradba_dbctl.sh stop --timeout 300

# Start and open all PDBs
oradba_dbctl.sh start ORCL --open-pdbs

# Force stop without confirmation
oradba_dbctl.sh stop --force

# Check status
oradba_dbctl.sh status

Behavior

Startup:

1. Checks oratab for databases with :Y flag (unless explicit SIDs provided)
2. For each database:
3. Sources environment via oraenv.sh
4. Checks if already running
5. Executes STARTUP;
6. Optionally opens PDBs with ALTER PLUGGABLE DATABASE ALL OPEN;

Shutdown:

1. For each database:
2. Sources environment via oraenv.sh
3. Checks if running
4. Attempts SHUTDOWN IMMEDIATE; with timeout
5. Escalates to SHUTDOWN ABORT; if timeout exceeded
6. Logs all operations

Justification:

When stopping ALL databases (no SIDs specified), the script shows a warning with the count of affected databases, asks for a justification (logged to file and console), and requires "yes" confirmation to proceed. Use --force to bypass in automation.

Environment Variables

Variable	Default	Description
ORADBA_SHUTDOWN_TIMEOUT	180	Default shutdown timeout in seconds
ORADBA_LOG	/var/log/oracle	Log directory
ORATAB	/etc/oratab	Path to oratab file

Listener Control

oradba_lsnrctl.sh manages Oracle listeners across different Oracle homes.

Features

• Supports multiple listeners across homes
• Defaults to LISTENER from first Oracle home in oratab
• Explicit listener name specification
• Discovers running listeners for status command
• Cross-home listener management

Usage and Options

oradba_lsnrctl.sh {start|stop|restart|status} [OPTIONS] [LISTENER1 ...]

Option	Long Form	Description
-f	--force	Skip confirmation prompts
-h	--help	Show help message

Key Examples

# Start default LISTENER
oradba_lsnrctl.sh start

# Start specific listeners
oradba_lsnrctl.sh start LISTENER LISTENER_ORCL

# Stop all listeners with force
oradba_lsnrctl.sh stop --force

# Show status of all running listeners
oradba_lsnrctl.sh status

Environment Variables

Variable	Default	Description
ORADBA_LOG	/var/log/oracle	Log directory
ORATAB	/etc/oratab	Path to oratab file
TNS_ADMIN	—	TNS configuration directory

Service Orchestration

oradba_services.sh orchestrates startup and shutdown of databases and listeners in configured order.

Features

• Configurable startup/shutdown order
• Default: start listeners first, stop databases first
• Delegates to oradba_dbctl.sh and oradba_lsnrctl.sh
• Configuration via oradba_services.conf
• Pass-through options to underlying scripts
• Unified status reporting

Usage and Options

oradba_services.sh {start|stop|restart|status} [OPTIONS]

Option	Long Form	Description
-f	--force	Skip confirmation prompts
-c	--config FILE	Use alternate configuration file
-h	--help	Show help message

Configuration File

Located at ${ORADBA_BASE}/etc/oradba_services.conf:

# Startup order (default: listener,database)
STARTUP_ORDER="listener,database"

# Shutdown order (default: database,listener)
SHUTDOWN_ORDER="database,listener"

# Specific databases (empty = all with :Y flag)
SPECIFIC_DBS=""

# Specific listeners (empty = default LISTENER)
SPECIFIC_LISTENERS=""

# Database options
DB_OPTIONS=""

# Listener options
LSNR_OPTIONS=""

Configuration Examples

Example 1: Production with PDBs

STARTUP_ORDER="listener,database"
SHUTDOWN_ORDER="database,listener"
SPECIFIC_DBS="PRODDB"
SPECIFIC_LISTENERS="LISTENER_PROD"
DB_OPTIONS="--open-pdbs --timeout 300"

Example 2: Development (all databases)

STARTUP_ORDER="listener,database"
SHUTDOWN_ORDER="database,listener"
SPECIFIC_DBS=""
SPECIFIC_LISTENERS=""
DB_OPTIONS="--open-pdbs"

Example 3: Multiple databases with custom timeout

STARTUP_ORDER="listener,database"
SHUTDOWN_ORDER="database,listener"
SPECIFIC_DBS="ORCL CDB1 TESTDB"
SPECIFIC_LISTENERS="LISTENER"
DB_OPTIONS="--timeout 240"

Log Management

OraDBA provides automated log rotation through logrotate integration. Proper log management addresses disk space growth, system performance, compliance retention requirements (PCI-DSS, HIPAA, SOX, GDPR), security audit trails, and historical troubleshooting data.

Prerequisites: logrotate installed (standard on most Linux distributions), root access for system-wide configuration.

Quick Start

# Install all logrotate templates (requires root)
sudo oradba_logrotate.sh --install

# Test configuration without rotating logs
oradba_logrotate.sh --test

# Force rotation for testing (requires root)
sudo oradba_logrotate.sh --force

# User-mode setup (no root required)
oradba_logrotate.sh --install-user
oradba_logrotate.sh --run-user

Log Templates

Five logrotate templates are provided in src/templates/logrotate/:

Template	Purpose	Default Rotation	Retention
oradba.logrotate	OraDBA system logs	Monthly/Weekly	12 months/8 weeks
oracle-alert.logrotate	Database alert logs	Daily	30 days
oracle-trace.logrotate	Trace files cleanup	Weekly (maxage)	30 days
oracle-audit.logrotate	Audit logs	Weekly	90 days
oracle-listener.logrotate	Listener logs	Daily	30 days

Logrotate Templates

Each template includes OraDBA headers, inline documentation, safe defaults, gzip compression, missingok/notifempty error handling, and appropriate create modes or copytruncate for active logs.

User-Mode vs System-Wide

For environments where root access is restricted, OraDBA supports non-root logrotate operation:

Aspect	User-Mode	System-Wide
Privileges	No root required	Requires root
Installation	~/.oradba/logrotate/	/etc/logrotate.d/
State files	~/.oradba/logrotate/state/	/var/lib/logrotate/
Execution	Manual or user crontab	System cron
Scope	User's Oracle logs	All system logs
Management	Self-service	System admin

Logrotate User-Mode vs System-Wide Comparison

User-mode setup:

# Initialize configuration
oradba_logrotate.sh --install-user

# Review and test
oradba_logrotate.sh --test

# Automate with cron
oradba_logrotate.sh --cron
# Then: crontab -e → paste generated entry

Compliance Requirements

The oracle-audit.logrotate template manages audit trails. Default retention is 90 days; adjust rotate values to meet regulatory requirements:

Standard	Retention Period	Notes
PCI-DSS	1 year minimum	Card payment data
HIPAA	6 years minimum	Healthcare data
SOX	7 years minimum	Financial records
GDPR	Varies by data	EU privacy law

Example PCI-DSS configuration:

# Edit /etc/logrotate.d/oracle-audit
rotate 365  # Daily rotation = 365 days

Example HIPAA configuration:

weekly
rotate 312  # 6 years * 52 weeks

System Integration

systemd Service

Copy and enable the systemd unit file:

sudo cp src/templates/systemd/oradba.service /etc/systemd/system/
sudo systemctl daemon-reload
sudo systemctl enable oradba.service
sudo systemctl start oradba.service

Manage the service:

sudo systemctl start oradba.service
sudo systemctl stop oradba.service
sudo systemctl restart oradba.service
sudo systemctl status oradba.service
sudo journalctl -u oradba.service -f

init.d/chkconfig Service

Copy the init.d script:

sudo cp src/templates/init.d/oradba /etc/init.d/
sudo chmod +x /etc/init.d/oradba

Red Hat/CentOS:

sudo chkconfig --add oradba
sudo chkconfig oradba on
sudo service oradba start

Debian/Ubuntu:

sudo update-rc.d oradba defaults
sudo update-rc.d oradba enable
sudo service oradba start

Service Log Locations and Format

All service management operations are logged to:

Log File	Contents
${ORADBA_LOG}/oradba_dbctl.log	Database operations
${ORADBA_LOG}/oradba_lsnrctl.log	Listener operations
${ORADBA_LOG}/oradba_services.log	Combined orchestration
${ORADBA_LOG}/oradba_services_root.log	Root wrapper operations

Service Management Log File Locations

Log format:

[2026-01-01 14:30:45] [INFO] Starting database ORCL...
[2026-01-01 14:30:50] [INFO] Database ORCL started successfully

Output to console uses color: INFO=green, WARN=yellow, ERROR=red.

Monitoring Integration

Monitor logrotate execution and log growth:

# Check logrotate status
ls -lrt /var/lib/logrotate/status
grep logrotate /var/log/messages

# Monitor log directory sizes
ORACLE_BASE=${ORACLE_BASE:-/opt/oracle}
THRESHOLD_MB=5000
for logdir in "$ORACLE_BASE"/diag/rdbms/*/*/trace; do
    [ -d "$logdir" ] || continue
    SIZE_MB=$(du -sm "$logdir" | awk '{print $1}')
    [ "$SIZE_MB" -gt "$THRESHOLD_MB" ] && echo "WARNING: $logdir is ${SIZE_MB}MB"
done

Alert if logrotate has not run recently:

LAST_RUN=$(stat -c %Y /var/lib/logrotate/status 2>/dev/null || echo 0)
AGE=$(( $(date +%s) - LAST_RUN ))
[ "$AGE" -gt 129600 ] && echo "WARNING: Logrotate hasn't run in $((AGE / 3600)) hours"

Troubleshooting

Database Won't Start

# Check environment
oraenv.sh ORCL
env | grep ORA

# Check if already running
ps -ef | grep pmon

# Manual startup to see error output
sqlplus / as sysdba <<EOF
STARTUP;
EXIT;
EOF

# Review script logs
tail -50 /var/log/oracle/oradba_dbctl.log
grep ERROR /var/log/oracle/oradba_dbctl.log

Listener Won't Start

# Check TNS configuration
ls -la $TNS_ADMIN

# Check if already running
ps -ef | grep tnslsnr

# Manual start
lsnrctl start LISTENER

# Review script logs
tail -50 /var/log/oracle/oradba_lsnrctl.log

Timeout Too Short

# Increase timeout for a single operation
oradba_dbctl.sh stop --timeout 600 ORCL

# Set permanently via environment
export ORADBA_SHUTDOWN_TIMEOUT=600

# Or set in services config
# Add to ${ORADBA_BASE}/etc/oradba_services.conf:
# DB_OPTIONS="--timeout 600"

Service Fails to Start at Boot

systemd:

sudo systemctl status oradba.service
sudo journalctl -xe -u oradba.service
systemctl list-dependencies oradba.service

init.d:

chkconfig --list oradba
sudo /etc/init.d/oradba start
tail -50 /var/log/oracle/oradba_services_root.log

Log Rotation Not Working

Symptom: Logs continue growing despite logrotate configuration.

# Test configuration
oradba_logrotate.sh --test --template oracle-alert

# Check for syntax errors
sudo logrotate -d /etc/logrotate.d/oracle-alert

# Debug in verbose mode
sudo logrotate -d -v /etc/logrotate.d/oracle-alert

Common causes: incorrect file paths or wildcards, permission issues, SELinux denials, syntax errors in configuration.

Permission Errors in Log Rotation

Symptom: permission denied errors in logrotate output.

# Verify file ownership
ls -la $ORACLE_BASE/diag/rdbms/*/*/trace/

# Logrotate configs must be root-owned
sudo chown root:root /etc/logrotate.d/oracle-*
sudo chmod 644 /etc/logrotate.d/oracle-*

# Log directories must be oracle-owned
chown -R oracle:oinstall $ORACLE_BASE/diag/

For SELinux environments:

sudo ausearch -m avc -ts recent | grep logrotate
sudo audit2allow -a -M logrotate_oracle
sudo semodule -i logrotate_oracle.pp

Database Can't Write to Alert Log After Rotation

Cause: Using create instead of copytruncate for active (open) log files.

Solution: Always use copytruncate for logs held open by the database process:

/path/to/alert_*.log {
    copytruncate  # REQUIRED for active logs
    daily
    rotate 30
    compress
    missingok
}

Missing oratab Entry

• Error is raised if an explicit SID is provided but not found in oratab
• Warning is issued (not an error) when auto-detecting from the :Y flag
• Listener already running is logged as INFO and is not treated as a failure

See Also

• Configuration
• Environment Management
• Troubleshooting

Navigation

Previous: SQL*Net Configuration Next: Troubleshooting