Log-based metrics in GCP - Mini Project: Build & Apply

Create a Log-based Metric in Google Cloud

📖 Scenario: You are managing a Google Cloud project that runs several applications. You want to monitor how many error logs are generated to quickly detect issues.

🎯 Goal: Create a log-based metric in Google Cloud that counts the number of error logs with severity ERROR.

📋 What You'll Learn

Create a log-based metric resource with the exact name error_count_metric.

Use the filter to select logs with severity=ERROR.

Set the metric type to counter to count occurrences.

Use the Google Cloud Logging API resource format.

💡 Why This Matters

🌍 Real World

Log-based metrics help monitor application health by counting specific log entries like errors or warnings.

💼 Career

Cloud engineers and DevOps professionals use log-based metrics to set up alerts and dashboards for system reliability.

1

Define the log-based metric resource

Create a dictionary called log_metric with keys name set to "error_count_metric" and filter set to "severity=\"ERROR\"".

GCP

# Create the log_metric dictionary with name and filter
# Your code here

Need a hint?

Use a Python dictionary with keys name and filter.

2

Add the metric descriptor type

Add a key metricDescriptor to log_metric with a nested dictionary containing metricKind set to "DELTA" and valueType set to "INT64".

GCP

log_metric = {
    "name": "error_count_metric",
    "filter": "severity=\"ERROR\""
}
# Add metricDescriptor with metricKind and valueType
# Your code here

Need a hint?

The metric kind DELTA counts changes, and value type INT64 means integer values.

3

Add a description for the metric

Add a key description to log_metric with the value "Counts the number of error logs".

GCP

log_metric = {
    "name": "error_count_metric",
    "filter": "severity=\"ERROR\"",
    "metricDescriptor": {
        "metricKind": "DELTA",
        "valueType": "INT64"
    }
}
# Add description key with the text "Counts the number of error logs"
# Your code here

Need a hint?

The description helps explain what the metric measures.

4

Complete the log-based metric configuration

Add a key labelExtractors to log_metric with an empty dictionary {} to complete the configuration.

GCP

log_metric = {
    "name": "error_count_metric",
    "filter": "severity=\"ERROR\"",
    "metricDescriptor": {
        "metricKind": "DELTA",
        "valueType": "INT64"
    },
    "description": "Counts the number of error logs"
}
# Add labelExtractors key with an empty dictionary
# Your code here

Need a hint?

Label extractors are optional but required as an empty dictionary to complete the metric resource.