docker - Provide deeper swarm nodes-related metrics

Question

I would like to be able to be able to monitor more accurately the manager nodes within a swarm cluster. I expect (eg) such a metrics :

node_swarm{instance="10.0.0.14:9100",job="node",node_id="123",node_name="node_7",node_availabilty="drain",node_manager_status="Reachable",node_status="unreachable"} 1  
node_swarm{instance="10.0.0.145:9100",job="node",node_id="456",node_name="node_8",node_availabilty="Active",node_manager_status="Unreachable",node_status="down"} 1

... if it could also provide the labels of the nodes, these would be the icing on the cake! ...

But for now, I didn't manage to provide such with node_exporter (or cAdvisor)... I may have missed some clue, but they don't sound like they provided such hints.

Do you have any idea on how I could provided such metrics?

I am quite sure I could code them from scratch (or submit a feature request), but I don't want to reinvent the Wheel if it already exists!

标签： dockerdocker-swarmmetricsexporterprometheus-node-exporter

Answer 1

您是否启用了dockerd 指标端点并查看它提供的端点？