This service communicates directly with Jira.
All teams are welcome to add metrics to chart/templates/configmap.yaml. Here is an excerpt:
data:
config.yaml: |
server: acme.atlassian.net
metrics:
-
query: 'project = ACME AND status changed to "waiting for support" after -1d'
metric_name: helpdesk_tickets_new_1d
metric_type: gauge
metric_help: 'new helpdesk tickets'
The server will be the same for all queries. If your metrics are not coming through, the user specified in the secret may require additional access to your tickets.
You can enter any JQL query in the field query, but the result's Total property (i.e. the number of results) will be submitted to Prometheus. It should make sense in relation to a given workday; often that will mean adding after -1d to your JQL expression, but at other times what you are tracking may be something like the total number of open tickets, in which case no time bracket is required.
The metric_name is the string you will query in Prometheus/Grafana. The usual recommendations apply. Be mindful that all Prometheus users will see all metrics, so it helps to use names that clearly identify a given metric as yours.
The metric_type could be counter, gauge or histogram. For business data, you may find gauge and histogram more useful than the counter collector, which works best with high-frequency inputs such as server response codes.
The metric_help field describes your metric in human-friendly terms.