Node Logs

Tool

Use the node_script tool to run this skill:

node_script: node="<node>", skill="node-logs", script="get-node-logs.sh", args="<args>"

Parameters

Required (one of):

Parameter	Description
--unit UNIT	Systemd unit name (e.g. containerd, kubelet). Use this or --file.
--file PATH	Log file path on the node (e.g. /var/log/messages). Use this or --unit.

Optional:

Parameter	Default	Description
--since DURATION	1h ago	Time range for journalctl (only with --unit), e.g. 30m ago, 2h ago, today
--grep PATTERN	—	Case-insensitive grep filter pattern
--tail N	200	Maximum number of output lines

Examples

Check containerd logs on a node:

node_script: node="node-1", skill="node-logs", script="get-node-logs.sh", args="--unit containerd --tail 50"

Search containerd logs for a specific image:

node_script: node="node-1", skill="node-logs", script="get-node-logs.sh", args='--unit containerd --grep "myregistry.com/myapp" --since "2h ago"'

Check kubelet logs for errors:

node_script: node="node-1", skill="node-logs", script="get-node-logs.sh", args='--unit kubelet --grep "error" --since "30m ago"'

Read a log file on the node:

node_script: node="node-1", skill="node-logs", script="get-node-logs.sh", args='--file /var/log/messages --grep "containerd" --tail 100'

Use Cases

Diagnose container runtime issues

Check containerd or cri-o logs when pods fail to start, images fail to pull, or containers crash unexpectedly.

Investigate kubelet problems

Check kubelet logs for node-level issues like pod evictions, volume mount failures, or resource pressure.

Check system logs

Read /var/log/messages, /var/log/syslog, or other log files for kernel or system-level issues affecting the node.