Oracle OS

Oracle OS

Purpose

Monitoring these operating system metrics is crucial for understanding the overall health and performance of your Oracle
Oracle OS monitor connects directly to the Oracle Database and uses system views to retrieve real-time operating system metrics

Monitored item

Available metrics to monitor:

CPU Usage (%): The percentage of CPU resources being consumed by the system
Memory Usage (%): The percentage of physical memory (RAM) being used on the server
Memory Usage (MB): The amount of memory being used in megabytes
Disk Usage (%): The percentage of disk space used on the database server
Disk Used (MB): The amount of disk space used in megabytes
Network In (bytes/sec): The rate of incoming network traffic in bytes per second
Network Out (bytes/sec): The rate of outgoing network traffic in bytes per second
Load Average (active sessions): The average system load based on the number of active sessions
Oracle Process Count: The number of Oracle background and user processes currently running
Oracle Memory Usage (MB): The amount of memory specifically consumed by Oracle processes (SGA, PGA, …)
Oracle Active Connections: The number of active database connections
System Load (Oracle sessions): The overall system load as measured by Oracle session count

Configuration hints

Create a new rule and select the OS metric to monitor from the “OS Feature” dropdown list.
Then set the threshold for the metric value. You have two options:
- Simple threshold: Enter a numeric value (e.g., “90” for 90%) and specify the severity level in the “Severity” field. An alarm will be triggered when the metric exceeds this value.
- Multi-threshold syntax: Use advanced syntax to define multiple threshold levels with different severities:
  - G2W:80 (Send a WARNING alert if value goes above 80 - Green To Warning)<br/>
  - W2M:95 (Send a MAJOR alert if value goes above 95 - Warning To Major)<br/><br/>
Example configurations:
- To monitor CPU usage and alert when it exceeds 80%: Select “CPU Usage (%)” and set threshold to “80” with severity “WARNING”
- To monitor disk space with multiple levels: Select “Disk Usage (%)” and set threshold to “G2W:75;W2M:90” to get a warning at 75% and major alarm at 90%
- To track Oracle connections: Select “Oracle Active Connections” and set threshold to “100” to alert when connection count exceeds 100
You can create as many rules as necessary and set multiple threshold levels for each metric. Each rule monitors a specific metric independently.
Use the “Auto clear” option to automatically clear alarms when the metric value returns to normal (below threshold).
The “Alarm tag” field allows you to customize alarm messages with context-specific information, making it easier to identify the source and nature of alerts.

Surveillance table

Parameter	Description
Active	To enable/disable a rule, unchecked rules will not be evaluated
OS Feature	Select the OS metric to monitor from the dropdown list
Threshold	The threshold value for the metric. If you set a simple number (e.g., “90”), the severity field will be used to define alarm severity. You can also use multi-threshold syntax (e.g., “G2W:80;W2M:95”) for multiple alert levels
Severity	If not DISABLED, send an alarm with this severity level when the threshold is exceeded. This field is ignored when using multi-threshold syntax
Auto clear	If checked, the alarm will be cleared automatically as soon as the metric value returns below the threshold. Recommended for dynamic metrics like CPU and memory usage
Alarm tag	This field allows you to add custom text within the alarm message, the %MSG% variable will contain the actual generated message and can be used such as: “my_prefix %MSG% my_suffix”. By default, the tag will be used as a prefix to the alarm message
Alarm	To enable or disable alarm generation. When unchecked, no alarms will be sent even if thresholds are exceeded
Metric	To enable or disable metric collection and forwarding to the monitoring platform. Metrics can be used for trending and historical analysis
Report	To enable or disable report mode for this rule (for scheduled report generation)

Generated metrics

The following metrics are generated by this monitor and can be forwarded to your monitoring platform:

metricId	metricUnit	metricTarget	metricDescription
ORACLE_OS_CPU_USAGE_PERCENT	%	OS	CPU usage percentage
ORACLE_OS_MEMORY_USAGE_PERCENT	%	OS	Memory usage percentage
ORACLE_OS_MEMORY_USAGE_MB	MB	OS	Memory usage in megabytes
ORACLE_OS_DISK_USAGE_PERCENT	%	OS	Disk usage percentage
ORACLE_OS_DISK_USED_MB	MB	OS	Disk used space in megabytes
ORACLE_OS_NETWORK_IN_BYTES_PER_SEC	bytes/sec	OS	Network incoming traffic per second (data received)
ORACLE_OS_NETWORK_OUT_BYTES_PER_SEC	bytes/sec	OS	Network outgoing traffic per second (data sent)
ORACLE_OS_LOAD_AVERAGE_SESSIONS	count	OS	System load average based on active database sessions
ORACLE_OS_PROCESS_COUNT	count	OS	Number of Oracle background and user processes running
ORACLE_OS_MEMORY_USAGE_MB	MB	OS	Memory consumed by Oracle processes (SGA, PGA, etc.) in megabytes
ORACLE_OS_CONNECTIONS_COUNT	count	OS	Number of active connections
ORACLE_OS_SYSTEM_LOAD_SESSIONS	count	OS	System load based on total Oracle session count

Best practices

CPU monitoring: Set thresholds at 75% (warning) and 90% (critical) to get early warnings before CPU exhaustion
Memory monitoring: Monitor both percentage and absolute values to understand memory consumption patterns
Disk space: Set alerts at 80% to have time to free up space or expand storage before reaching critical levels
Connection monitoring: Know your connection pool limits and set alerts accordingly (typically 80% of max_connections)
Network monitoring: Baseline your normal network traffic and alert on unusual spikes that could indicate issues
Enable metrics: Always enable metric collection for trending and capacity planning, even if you don't need immediate alarms

Table of Contents

Oracle OS

Purpose

Monitored item

Configuration hints

Surveillance table

Generated metrics

Best practices