Today, we’re excited to announce the ability for Service Connector Hub to create dimensions from Oracle Cloud Infrastructure (OCI) Logging data when sending logs as metrics to the OCI Monitoring service.

Until now, Service Connector Hub has allowed you to specify logging queries. Whenever they find resultant logs, such as error cases, the number of log records are then converted into data points that it can plot on charts or alarm on using the OCI Monitoring service. With dimensions support, we’re allowing you to attach more metadata to your generated metrics and get finer control over your monitoring dashboards and alarms.

Metric data and dimensions

Metric data consists of data points. A data point is simply a collection of properties that are records of observations taken within a system. Each data point consists of the following elements:

  • Metric name: The name of the observation we’re recording, such as CPU usage, network throughput, and number of connections.

  • Metric value: The observed value of the observation we made. When plotting a metric on a chart, these values are represented on the y-axis.

  • Timestamp: The time at which the observation was made. When plotting a metric on a chart, these values are represented on the x-axis.

  • Dimensions: Metadata about which part of the system the observation was made in. Can also be thought of as tags or labels.

A graphic explaining the four parts of a datapoint.

Dimensions, also known as tags, labels, or filters, allow you to attach other metadata to the metric value. They provide the difference between telling your friend visiting from New York that you live in Seattle versus your mailing address in Seattle. In the second example, you can meet your friend, while in the first example, your friend can wander around Seattle trying to find where you live. So, dimensions are like the house number, street address, and zip code, which help pinpoint your exact location in Seattle.

Quick walkthrough

Start by selecting “Logging” as the source and “Monitoring” as the destination.

A screeshot of the Configure Service Connector screen with Logging as the chosen source and Monitoring as the target.

Make your selections for the log sources that you want to query. In the following screenshot, I’m checking for an “anomaly detected” event log.

A screenshot of the Configure Source screen with details with details filled in.

In the Configure Target section, you can see the new ability to add dimensions.

A screenshot of the Configure Target screen with a red arrow pointing to the Add Dimensions button.

This selection brings up the following panel, which shows the last six log lines based on your log source selection.

A screenshot of the Add Dimension window with details filled in.

Opening one of the log rows lets you select the exact property that you want to choose as a dimension. Selecting the property adds the exact jmespath needed to extract the dimension value from the log.

A screenshot of the Select Path screen with numbered red arrows pointing to the menu expansion carat, availability domain, and dimension name.

With dynamically picking dimensions from your log data, you can also supply static dimensions for data that differentiates a metric but is not present in the log.

A screenshot of the Static Values screen with details filled in.

And you’re done! You just set up a metric with dimensions. When the metric starts flowing to the OCI Monitoring service, you can plot the chart or create an alarm using the existing interfaces, such as Metrics Explorer.

A screenshot of the Metrics Explorer with the dimensions noted with red numbers.

Getting started

To get started, visit one of the following documentation links: