From a blank machine (CentOS 7) everything provisions OK until datasources:
Notice: /Stage[main]/Grafana::Service/Service[grafana-server]/ensure: ensure changed 'stopped' to 'running'
Info: /Stage[main]/Grafana::Service/Service[grafana-server]: Unscheduling refresh on Service[grafana-server]
Error: /Stage[main]/Profile_grafana/Grafana_datasource[elasticsearch]: Could not evaluate: Network is unreachable - connect(2)
Error: /Stage[main]/Profile_grafana/Grafana_datasource[graphite]: Could not evaluate: Network is unreachable - connect(2)
There's not enough time for the service to actually be ready to accept connections. Executing the same 'Puppet run' again succeeds as expected. The send_request would ideally have some 'try once per second for five seconds' logic to make it more robust.
From a blank machine (CentOS 7) everything provisions OK until datasources:
There's not enough time for the service to actually be ready to accept connections. Executing the same 'Puppet run' again succeeds as expected. The
send_request
would ideally have some 'try once per second for five seconds' logic to make it more robust.