Hosts (Edge Node)

Edge Node observability for Hosts provides visibility into the health and performance of the Edge Node hosts by collecting logs and metrics from the Edge Node and shipping them to a centralized location for viewing and analysis. The following data is collected for every host of the Edge Node:

• Metrics are collected via Telegraf telemetry collector:

  • Configured with a selected set of plugins focused on bare-metal metrics.

  • Set of plugins can be extended via Edge Infrastructure Manager’s Telemetry Control subsystem.

• Logs are collected via Fluentbit log collector instances:

  • provisioning logs - via separate instance deployed as part of HookOS (review Overview and its subsections for an overview of provisioning).

  • agent logs - logs from the agents running on the Edge Node.

  • host logs - logs from the host system.

  • health-check logs - additional instance monitoring the health of other services.

Note

Platform Observability Agent also forwards logs from a cluster if one is deployed on the Edge Node. Refer to Clusters (Edge Node) for more details.

Bare-metal host logs and metrics are collected and shipped out of the Edge Node via Platform Observability Agent. Optional additional data along with collection interval can be controlled via Platform Telemetry Agent. For additional details, refer to:

• Edge Node Platform Observability Agent

• Edge Node Platform Telemetry Agent

Note

All data originating from a single Edge Node is associated with a single project (tenant). Refer to Multitenancy section for more details on how multitenancy in observability is handled.