Analyze GCP Error Logs
Automatically fetch recent Google Cloud Platform error logs, group them by root cause using AI, and send a detailed HTML report to your team daily.
Introduction
This workflow automatically analyzes Google Cloud Platform (GCP) log entries for errors over the past day and generates a detailed HTML report. It leverages the GCP logging API, authenticates with a service account, and uses an AI model to cluster error logs by root cause and provide actionable insights.
It performs the following tasks:
- Runs on a daily schedule at 9:20 AM Berlin time.
- Creates a JSON Web Token (JWT) to authenticate with the Google OAuth 2.0 system.
- Requests an access token using the generated JWT.
- Queries the GCP Logging API for error log entries from the past 24 hours filtered by severity.
- Sends the log entries to an AI model that analyzes and clusters errors by root cause, summarizes the data, identifies affected services, and offers recommendations.
- Processes the structured response from the AI and generates an HTML report with detailed tables and links to logs.
- Outputs a structured, easy-to-read report for developers and operations teams to identify and respond to issues.
What you need
- A GCP project with logging enabled.
- A service account JSON key with permissions to read logs and generate OAuth tokens.
- The private key and client email of the service account available as input.
- Access to the Needle platform with the ability to create scheduled workflows.
- An active Needle AI integration for the log analysis step.
How the flow works
Here is a breakdown of the nodes used in this workflow:
| Step | Component | Description |
|---|---|---|
| 1 | Scheduled trigger | Starts the workflow every day at 9:20 AM Europe/Berlin time. |
| 2 | JWT creation code | Parses your GCP service account key and builds a signed JWT for OAuth authentication. |
| 3 | OAuth token request | Sends the JWT to the Google OAuth token endpoint to get an access token. |
| 4 | Logs query request | Uses the access token to call the GCP Logging API and fetch error logs from the last 24 hours. |
| 5 | AI log analysis | Passes the raw error entries to an AI model to group logs, summarize patterns, and provide recommendations. |
| 6 | Report generation code | Takes the structured output along with raw stats to build an HTML report with error clusters and log links. |
| 7 | Sends the final HTML report directly via Gmail. |
AI Analysis Capabilities
The AI model is specially prompted to extract the following insights:
- Group logs by root cause.
- Summarize main error patterns.
- List affected services.
- Analyze error timing.
- Provide prioritized recommendations.
Output
At the end, you get a comprehensive HTML report detailing:
- Report date and total error count.
- Executive summary of main error patterns and affected services.
- A table of error clusters with descriptions, sample errors, affected services, and log links.
- Service-level error counts and top clusters.
- Timeline analysis describing error spikes.
- Prioritized recommendations for addressing the identified issues.
This report helps development and operations teams quickly understand error trends and root causes in their GCP environment.
Notes
- The service account provided must have sufficient permissions to read logs.
- The AI analysis depends on accurate and comprehensive logs; ensure your GCP logging configuration captures relevant error information.
- Be mindful of API quotas on GCP and Needle when running this workflow daily.
- Customize the AI prompt or report formatting to fit your team specific needs.
Want to showcase your own workflows?
Become a Needle workflow partner and turn your expertise into recurring revenue.