What it does
Oracle Cloud Infrastructure Monitoring is a metrics and alarms service for workloads in an OCI tenancy. It presents ready-made resource metrics, accepts custom application measurements, aggregates query results, and evaluates alarms that can notify people through OCI Notifications or Streams. Teams can use the Console, CLI, API, and Monitoring Query Language to inspect service health and performance. Its strongest fit is a team already operating on OCI that wants resource and application metrics in the same monitoring service.
Start with an OCI metric
The service automatically receives metrics from supported OCI resources. A practical first check is a compute instance's CPU utilization: choose the resource metric, the time range, and an aggregation such as a mean or maximum, then inspect the chart for changes that need attention. The Console also provides Metrics Explorer and supports Monitoring Query Language (MQL) for queries across resource groups.
For an application-specific signal, publish a custom metric through the Monitoring API, an SDK, the CLI, or an agent configuration. Give it an application-owned namespace and dimensions, and make sure the identity publishing it has the required IAM policy. Oracle reserves the oci_ and oracle_ namespace prefixes. Custom data can be sent as frequently as once per second, but the minimum aggregation interval is one minute; API requests also have data-point and timestamp limits.
Use alarms for detection
An alarm evaluates a metric query and sends a message to a configured destination when its trigger is met. Destinations include OCI Notifications topics and Streaming, so an alarm can page a team or feed an existing event workflow. Alarms are evaluated once per minute. They detect a condition and deliver its signal; an operator or connected workflow still needs to diagnose and correct the underlying resource problem.
All OCI tenancies receive an Always Free allowance of 500 million Monitoring ingestion datapoints and 1 billion retrieval datapoints. Higher usage may be billed; Oracle's price list is region- and currency-sensitive, so confirm current rates before estimating a busy tenancy. This is an OCI service, not a separate account-free monitor.
Strengths
- Built-in OCI resource metrics and custom application metrics can be queried through one service.
- Metric queries can drive alarms that send notifications to OCI Notifications or Streaming.
- The Console, CLI, API, SDKs, and MQL provide several ways to inspect or publish measurements.
Watch out for
- Custom metric publishing requires an appropriate IAM policy and must follow namespace, timestamp, aggregation, and request-size constraints.
- Alarm evaluations run once per minute, so this is not a sub-minute alerting path.
Best for
- OCI operators who want CPU, storage, network, or other supported resource metrics beside service alarms.
- Application teams exporting custom counters or latency measurements to OCI for charting and alerting.
Sources
- Monitoring | Oracle (Oct 26, 2026)
- Overview of Monitoring (Oct 26, 2026)
- Publishing Custom Metrics Using the API (Oct 26, 2026)
- Cloud Price List | Oracle (Oct 26, 2026)
- Always Free Resources (Oct 26, 2026)