Annotations can mark events on your Metrics dashboards based on data from a logging account. For example, an annotation can indicate when a particular error log or configuration change occured based on logs written to your main logging account.
Configure an annotation
Pick your dashboard
Before you begin, you'll need to make sure you have permissions to edit the dashboard.
If you are using any of the pre-configured dashboards provided by Logz.io, you'll first need to select the option to make it editable. (You can also click the gear in the top menu, then select Save as > Save).
Add a new annotation
Go to the dashboard's settings:
- Click the gear in the dashboard's top menu (upper right corner).
- Select Annotations > New.
- The annotation form will appear. Fill in the form, starting with the General section.
- Select an informative Name. The name should describe the significance of the event.
- Select a Data source from your logging accounts. This is the relevant OpenSearch Dashboards account you want to correlate with your metrics.
Decide which logs should trigger the annotation
You'll be using a OpenSearch Dashboards search query to select the logs you want to mark on your dashboard. They will appear as vertical markers on the dashboard's timeline to indicate when they were sent.
Type in an OpenSearch Dashboards query in Lucene syntax. It's best if you double-check your query results in OpenSearch Dashboards to make sure it is specific enough.
You can use Metrics interface variables, on condition that the value of the variable actually appears in the target log. For example, the query
logtype:configuration AND pod.name:$pod would work assuming the fields
pod.name exist in the log.
Enrich the annotation
You can make annotations more informative by specifying the fields of interest. These should be fields that contain the crucial information from the log. Their values will be shown when you hover over the annotation in the dashboard.
Text - Type in the log field that is of most interest. The field's value will be shown when hovering over the marker. In the example below, the
messagefield was selected.
Tags - Type in one or more log fields. Their values will appear as tags when you hover over annotations in the dashboard. In the example below, the
servicefield was selected.
Save the annotation & dashboard
When you're finished, save the annotation to see it in action.
Once you're satisfied with the annotation, save the changes to the dashboard. (A prompt will make sure you don't forget this part.)
As an example, we'll go through the steps of adding an annotation to flag every time there's a configuration change in a Kubernetes cluster.
This can make it easier to assess whether a particular configuration change caused problems in a cluster.
After selecting the relevant log account, define a query that will return only Kubernetes configuration change logs. Here's what the query might look like:
loglevel:Configuration AND namespace:"kube-system"
Whenever a log matching this query is returned, it will appear as a vertical annotation on the dashboard. The hover message for the annotation will show the log's
message field with the
service field tag.
Here's what the configuration will look like:
And here's what the annotation will actually look like in the dashboard.