cloudwatch_exporter
cloudwatch_exporter copied to clipboard
I can't see metrics values in prometheus
Hello everyone,
Thank you for this amazing exporter. Probably I am doing something wrong. I need to export some metrics from AWS MSK. I used prometheus exporter helm chart from prometheus community repo.
My metrics config:
region: us-west-2 period_seconds: 240 metrics:
- aws_namespace: AWS/Kafka aws_metric_name: MemoryUsed aws_dimensions: [Cluster Name, Broker ID]
my service monitor config
serviceMonitor:
When set true then use a ServiceMonitor to configure scraping
enabled: true namespace: monitoring interval: 30s telemetryPath: /metrics timeout: 10s
I can see the metrics on cloudwatch exporter endpoint but I cant see in prometheus, the service monitor is enabled



Do the metrics show up in the graph view? Could this be due to the default delay_seconds
(10 minutes)? Unfortunately CloudWatch metrics only converge after a while, while Prometheus samples are immutable, so the exporter has to (usually) report data from some time in the past.
Same issue. The cloudwatch-exporter is deployed with prometheus as sub-chart. Deployed servicemonitor which gets picked up by prometheus servicediscovery, but (0 / 46 active targets). None of the exposed metrics from cloudwatch-exporter ever show up in prometheus. How do we get cloudwatch-exporter metrics to show up in prometheus?
@mr-coveros if the exporter isn't being discovered, something is not working at the discovery stage, which is different from @rubenssoto's problem. Looking at this troubleshooting document, a common issue is port name mismatches.
Since the metrics show up in the exporter, this isn't an exporter issue, so I'll close this one to avoid distracting others from unrelated configuration issues.