Sign in Subscribe

Topic

Metrics

A collection of 2 issues

Tracing and Monitoring IPFS Cluster

IPFS Cluster supports exposing a Prometheus endpoint for metric-scraping as well as submitting trace information to Jaeger. These are configured in the observations section of the configuration and can be enabled from there or by starting a cluster peer with: $ ipfs-cluster-service daemon --stats --tracing Development setup for tracing and metrics

IPFS Cluster Metrics

Cluster peers run a “monitor” component which is in charge of: * Distributing arbitrary metrics from the current peer to all the others. Metrics have an associated expiry date. * Detecting peer-health events and triggering actions * In crdt-mode, the peerset is defined as the list of peers for which we have received