This article describes how to integrate Anodot and Grafana, and includes:
- Installation and Configuration
- Using Anodot-based Panels
- Specific Scenario Configurations
- Using Variables for Metric Dimensions
Installation and Configuration
Integrating with Grafana enables you to view Anodot information in your Grafana dashboards. In order to correctly use the integration, you first need to add the following two Anodot Grafana Plugins in your Grafana installation:
- Anodot Grafana Data Source - Enables a secure connection between Grafana and your Anodot account.
- Anodot Grafana Panel - Enables the visualisation of the the Anodot data in Grafana.
You need both of these plugins in order to consume Anodot data in Grafana. The fastest way is to get the plugins from the official Grafana plugin directory:
Alternatively, you can self-build the plugins from GitHub:
- Data Source - https://github.com/anodot/grafana-datasource
- Panel - https://github.com/anodot/grafana-panel
Once the plugins are installed in your Grafana setup you need to configure the data plugin to point to your Anodot account.
- From the Data Sources menu in Grafana, choose the Anodot data source, and in the Settings tab enter your Anodot API URL and Anodot API Token, as shown below. For more information on how to obtain these, see here.
Please note - If you upgraded from version 3.x to 4.x of the data source plugin, you may need to update the URL manually adding "/" at the end.
- Click Save & Test to verify that you've entered the correct token for your account.
Using Anodot-based Panels
Once you've configured the data source, you can start creating Anodot-based panels. Choosing the data source you've just defined in the Query tab, you will that there are several scenarios supported:
- Alerts - Shows a list of alerts from Anodot
- Metrics - Shows a list of metrics from Anodot
- Anomalies - Shows a list of anomalies from Anodot
- Topology Map - Shows a topological map of alerts for metrics that support this display
After choosing a scenario, the panel changes to reflect the possible parameters for each scenario, as shown below. Configuring the parameters for each panel enables you to get the relevant data (see the detailed explanations in the following section). Note that setting the time frame for the panel is done in the top right corner of the panel, as for any Grafana panel.
Specific Scenario Configurations
The Alerts scenario enables you to build a panel based on alerts that Anodot has detected. You can filter these alerts based on the following parameters:
- Recipients / Channels
- Alert Type - Anomaly / Static / No Data
- Severity - Critical / High / Medium / Low
- Feedback - Good Catch / Not Interesting / No Feedback
- Acknowledge - All, Acknowledged, Not Acknowledged
Once you've defined the panel, the different columns are sortable (Started, Duration, Score, Name). Clicking on an entry in the Alerts list will open the alert in Anodot.
The Metrics scenario enables you to view metrics as they come from Anodot. After choosing this scenario, you must first choose a measure - meaning the Anodot measure you would like to view in Grafana. Choosing the measure populates the list for possible dimensions this measure has; you can then filter according to specific dimension values.
There are two additional parameters which you can set:
- Include baseline - enables you to view the Anodot baseline as the background to the metric (Highly recommended).
- Multiline Mode - combines all the metrics which meet the condition into one panel.
Coming soon: the ability to choose a function on top of the 'raw' metric.
The Anomalies scenario enables you to view anomalies detected by Anodot (similar to the Anomalies page in Anodot). Similar to the Metrics scenario described above, you first need to choose a measure for which you would like to get anomalies. As there are many different types of anomalies, you can filter according to the following parameters:
- Anomaly Duration
- Anomaly Score
- Delta Type
- Delta Value
- Time Scale
- Sort order
- Open Anomalies Only
The resulting panel shows the list of filtered anomalies according to the Anodot color coding (light blue for the baseline, dark blue for the actual data, and orange for the anomaly).
The Topology Map scenario enables the showing of anomalies that are part of a specific topology. For example, network topology, Kubernetes cluster, etc.
The main requirement for measures to be displayed using this scenario is that there is a dimension that can serve as the source vertex and a dimension that can serve as the destination vertex. Anodot will draw edges between these vertices. When Anodot identifies anomalies on these metrics, the relevant edges will be colored accordingly.