Google Cloud Logging and Monitoring

Cloud Logging (formerly Stackdriver Logging) is a fully managed log management service that lets you store, search, analyze, and export logs from your applications and GCP services.

Key Features:

• Ingest logs from Compute Engine, Cloud Run, GKE, App Engine, and custom apps.

• Powerful log filtering and querying with Log Explorer.

• Export logs to Cloud Storage, BigQuery, or Pub/Sub.

• Set up log-based alerts.

Cloud Monitoring (formerly Stackdriver Monitoring) lets you visualize metrics, build dashboards, and receive alerts based on the performance of your systems.

Key Features:

• Collect metrics from GCP services, AWS, and on-prem systems.

• Prebuilt and custom dashboards.

• Uptime checks and alerting policies.

• Integrates with Cloud Logging, Error Reporting, Trace, and Profiler.

Hands-On: Monitor a VM Instance

Let’s walk through a basic example of using Logging and Monitoring with a Compute Engine VM.

Step 1: Enable APIs

Make sure these APIs are enabled in your project:

gcloud services enable monitoring.googleapis.com logging.googleapis.com

---

Step 2: Create a VM with Logging Enabled

gcloud compute instances create monitored-vm \
  --zone=us-central1-a \
  --metadata=startup-script='#!/bin/bash
  sudo apt-get update
  sudo apt-get install -y apache2
  systemctl start apache2' \
  --scopes=https://www.googleapis.com/auth/logging.write,https://www.googleapis.com/auth/monitoring.write

This startup script installs and runs Apache — and sends system and service logs to Cloud Logging automatically.

Step 3: View Logs in Log Explorer

• Go to Cloud Console > Logging > Logs Explorer

• Use this query to see VM logs:

resource.type="gce_instance"
resource.labels.instance_id="YOUR_INSTANCE_ID"

Step 4: Create a Custom Dashboard

1. Go to Cloud Console > Monitoring > Dashboards

2. Click Create Dashboard

3. Add a line chart for the metric compute.googleapis.com/instance/cpu/utilization

4. Apply filters to your instance

Step 5: Set an Alerting Policy

1. Go to Monitoring > Alerting > Create Policy

2. Add a condition:

   • Metric: CPU Utilization

   • Trigger when CPU > 80% for 5 minutes

3. Add a notification channel (email, SMS, Slack, etc.)

4. Save the policy

Exporting Logs (Optional)

To archive or process logs:

gcloud logging sinks create my-logs-to-storage \
  storage.googleapis.com/YOUR_BUCKET_NAME \
  --log-filter='resource.type="gce_instance"'

Best Practices

• Use log-based metrics to turn log entries into monitoring data.

• Tag logs with custom labels to improve searchability.

• Keep alert rules meaningful — avoid alert fatigue.

• Use uptime checks for external endpoint monitoring.