Important log Files in Platform9 Managed Kubernetes

This document will list all the important log files that can be referred to troubleshoot issues with your cluster.

  • Hostagent log: /var/log/pf9/hostagent.log
  • This log file is created when you run the pf9ctl prep-node command. This log will help you identify the status of the node from a management plane perspective. All Platform9 related package download and installations are logged here. If you do not see your node in a Healthy state or observe any errors during the node prep, head over to this file to see where the failure occured.

Once you onboard a node successfully, in addition to the hostagent log, you will see the following directories and log files created under the /var/log/pf9 directory.

  1. Comms log directory: /var/log/pf9/comms/
  • The comms directory keeps a record of the logs for the pf9-comms component running on the nodes. This log records all the communication between the node and the management plane. If you observe your node in an Unhealthy state, this can be another location to observe if there are any network connectivity issues.
  1. Kube log directory: /var/log/pf9/kube/
  • The kube log directory logs the activities of the pf9-kube service running on your Kubernetes node. There is no reason to panic if you observe an inactive status in this pf9-kube log since it will be in a stopped state until the node is added to a cluster.
  • This file stores information about the installation of Kubernetes role on a node and the output of periodic status checks performed on the node. Consult this file on the node for more information if you are running into issues with attaching the node to the cluster or if the node is reported as ‘Unhealthy’ in the UI.
  • Once the node is added to a cluster, this log file records all the activities performed by the pf9-kube service. Some actions performed by the pf9-kube service include, generating certificates, setting up docker and etcd, confihuring the CNI, configuring kubelet service, and running health checks.
  1. Kubelet log directory: /var/log/pf9/kubelet/
  • The kubelet directory will be empty until the node is added to a Kubernetes cluster and assigned a master/worker role.
  • Once the node addition to cluster is completed, kubelet.INFO, kubelet.ERROR, and kubelet.WARNING files are created to log all the events like pod creation/failures, node status change.
  1. Muster logs: /var/log/pf9/muster.log
  • These are the log files created by the pf9-muster service. The muster log maintains historic data about the node statistics like the load average, memory usage, and network statistics.
  1. Nodelet logs: /var/log/pf9/nodelet.log
  • The pf9-nodeletd service logs here. This service helps keep a track of multiple scripts/phases run by the pf9-kube service to attach/detach a node from the cluster. The nodelet log also logs any failures if your Kubernetes node is in a NotReady state.
  1. Sidekick log directory: /var/log/pf9/sidekick/
  • The pf9-sidekick service is responsible for sending heartbeats to the management plane. if there are any network connectivity issues, the sidekick service will log those in its log file.
1 Like