Grafana Dashboard Integrations

Download this manual as a PDF file

Features

  • You can configure Zebrium to automatically add Root Cause (RCA) reports in Grafana. This allows you to see details of root cause on any Grafana dashboard.
  • This integration automatically adds log metrics in Grafana.
  • Each Zebrium RCA report includes a summary, a word cloud, and a set of log events showing symptoms and root cause, plus a link to the full report in the Zebrium user interface.
  • This means faster Mean Time to Resolution (MTTR) and less time manually hunting for root cause.

How It Works

The recommended mode of operation for observability dashboard integrations is to use the Zebrium Auto-Detect mode as an accurate mechanism for explaining the reason something went wrong. In this mode, you continue to use your existing rules, alerts and metrics as the primary source of problem detection. You can then review Zebrium RCA report findings directly in your Grafana Dashboards alongside other metrics to explain the reason behind problems you were alerted on.

Auto-Detect: View Root Cause Detections to your Grafana Dashboards

  1. Zebrium continuously monitors all application logs and uses unsupervised machine learning to find anomalous log patterns that indicate a problem. These are automatically turned into Root Cause reports highlighting details of any problems with over 95% accuracy.
  2. Root Cause report summaries and Log metrics are visible on your Grafana Dashboards using the Zebrium plugin.
  3. If you need to drill down further to look at correlated logs across your entire app, it’s just one click from your Dashboard.

Viewing Root Cause Detections in your Grafana Dashboards

Integration Overview

  1. Download Plugins from GitHub.
  2. Sign and Install Plugins.
  3. Create a secure access token in Zebrium for the Zebrium Datasource.
  4. Setup the Zebrium Datasource.
  5. Install the Zebrium Root Cause Finder on a dashboard.

Prerequisites

  • File and Web access to a Grafana Server running at minimum version 8.0. We recommend first downloading and installing a local copy of Grafana Server to test the setup procedures, before installing the plugins on your production Grafana instance. You can install a local copy of Grafana web server from here: https://grafana.com/grafana/download/8.3.6?pg=get&plcmt=selfmanaged-box1-cta1.

  • Grafana Toolkit installed on your system. To install, run the npm command:

    npm i @grafana/toolkit

  • Web access to your Zebrium instance.

Current Zebrium Plugins

  • Zebrium Root Cause Finder. Plots counts of all log events, error events and rare events overlaid with detections found by the Zebrium AI/ML engine.
  • Zebrium Data Source. Provides the API access layer for Zebrium panel plugins.

Integration Details

STEP 1: Download Plugins from GitHub

Use the following path to download the Zebrium Grafana plugins:

https://github.com/zebrium/grafana-plugin/releases

This is a private repository in the current phase. For access, please contact Zebrium at support@zebrium.com.

STEP 2: Sign and Install Plugins

  1. Stop your Grafana Server.

  2. Move the plugins you downloaded from GitHub under your Grafana Server plugins directory. For example, if your Grafana Server is installed on the path /Grafana-8.3.6, you wuld install the downloaded plugins in /Grafana-8.3.6/data/plugins.

  3. Make sure that the Grafana Toolkit is installed on this system. If it isn’t, run the npm command:

    npm i @grafana/toolkit

  4. Follow the Grafana signing instructions for signing a private plugin: https://grafana.com/docs/grafana/latest/developers/plugins/sign-a-plugin/.

  5. You must sign each plugin separately and you must sign them as a private plugin:

    • cd /<PATH TO GRAFANA SERVER>/data/plugins/ze-datasource and sign as a private plugin.
    • cd /<PATH TO GRAFANA SERVER>/data/plugins/ze-detections and sign as a private plugin.

  6. Restart your Grafana Server. For example, if your server is in /Grafana-8.3.6, switch to that directory and run ./bin/grafana-server web.

  7. Check the Grafana server logs to verify that the new plugins have been registered:

    Image of the Grafana plugin registered page

STEP 3: Create a Secure Access Token in Zebrium

  1. In the Zebrium user interface, go to the Access Tokens page (Settings () > Access Tokens).
  2. Click + Add Access Token button.
  3. Enter a Name for the token.
  4. Select Viewer for the Role.
  5. Select the Deployment for the access token.
  6. Click Add.
  7. Copy the Access Token that was just created and save it for use in STEP 4.

STEP 4: Set up the Zebrium Datasource

  1. Open the Grafana server in a web browser. If you have installed a local Grafana Server, it is normally found at http://localhost:3000.

  2. On the left toolbar, find the Configure menu (gear icon) and select data source.

  3. Click Add a Datasource.

  4. Scroll through the list of data sources and select Zebrium Data Source.

  5. On the Settings page, enter the API endpoint. This is the FQDN for your Zebrium instance where you generated the Access Token in STEP 3. For Zebrium SaaS, this will typically be: https://cloud.zebrium.com.

  6. Also on the Settings page, enter the access token you created from STEP 3.

  7. Click Save & test.

  8. Verify that you see the message Successfully connected to Zebrium. If you do not see it, review your entries above.

    Image of the Zebrium Data Source dialog

STEP 5: Install Zebrium Root Cause Finder on a Dashboard

  1. Open the Grafana server in a web browser. If you have installed a local Grafana Server, it is normally found at http://localhost:3000.

  2. On the left toolbar, find the Create menu (plus icon) and select dashboard.

  3. On this new dashboard, click Add an Empty Panel. The panel setup screen appears.

  4. In the upper right, open the list of visualizations and select the Zebrium Root Cause Finder plugin.

  5. In the query panel, select Zebrium Data Source.

  6. For API Name, enter detections.

  7. For Service Group, enter All to pull counts and detection data from all of your service groups. Alternatively, enter the name of a single service group to pull data from just that service group.

  8. Give your panel a name; we suggest Zebrium Root Cause Finder. Then save your Dashboard and give it a name.

    Image of the Zebrium Root Cause Finder panel

  9. Verify that the Zebrium Root Cause Finder visualization shows up in the panel you just added:

    Image of the Zebrium Root Cause Finder panel with word cloud