How to use azure_sd_configs to scrape azure VM's and VMSS's

21 views

Skip to first unread message

Kirti Ranjan Parida

unread,

Jun 30, 2020, 9:38:33 AM6/30/20

to Prometheus Users

I am trying to scrape VM's and VMSS's in azure using prometheus.For discovery I am using azure_sd_configs.Below is my configuration

- job_name: 'node_exporter'

azure_sd_configs:

- authentication_method: ManagedIdentity

subscription_id: <subscriptionID>

refresh_interval: 5m

port: 9100

Do I have to mention anything ?

I am getting below error

level=warn ts=2020-06-30T08:13:55.895Z caller=klog.go:86 component=k8s_client_runtime func=Warningf msg="/app/discovery/kubernetes/kubernetes.go:261: watch of *v1.Endpoints ended with: too old resource version: 1291352 (1292845)"

level=error ts=2020-06-30T08:15:28.938Z caller=refresh.go:98 component="discovery manager scrape" discovery=azure msg="Unable to refresh target groups" err="could not get virtual machines: could not list virtual machines: azure.BearerAuthorizer#WithAuthorization: Failed to refresh the Token for request to https://management.azure.com/subscriptions/<REMOVEDSUBSCRIPTION>/providers/Microsoft.Compute/virtualMachines?api-version=2018-10-01: StatusCode=404 -- Original Error: adal: Refresh request failed. Status Code = '404'. Response body: getting assigned identities for pod <PROMETHEUS POD NAME> in CREATED state failed after 16 attempts, retry duration [5]s. Error: <nil>\n"