-
Spike
-
Resolution: Won't Do
-
Normal
-
None
-
None
-
None
-
None
-
False
-
False
-
Undefined
-
Customers are looking to use the centralised prometheus for monitoring/alerting through the federation route, we need to evaluate this approach and see whether there are any issues. One configuration we are aware of is
apiVersion: monitoring.coreos.com/v1
kind: ServiceMonitor
metadata:
name: istio-federation
namespace: istio-system
spec:
selector:
matchLabels:
app: prometheus
endpoints:
- interval: 30s
scrapeTimeout: 30s
params:
'match[]':
- '{job="pilot"}'
- '{job="envoy-stats"}'
path: /federate
targetPort: 9090
honorLabels: true
metricRelabelings:
- sourceLabels: ["__name__"]
regex: 'workload:(.*)'
targetLabel: "__name__"
action: replace
Questions to be answered are:
- What should the scope of this support be?
- What features can we enable through prometheus federation?
- Are there any conflicts/issues/bugs we need to address? (e.g. version mismatches?)
- What is the difference to "Bring your own Prometheus (
OSSM-1352)"?
Timebox: 3 days
