common

type ContainerdConfig struct {
	// Use unix:///run/k3s/containerd/containerd.sock for RKE2/k3s.
	Endpoint string `json:"endpoint,omitempty"`
	// Use /var/lib/rancher/rke2/agent/etc/containerd/config.toml for RKE2.
	ConfigPath string `json:"config_path,omitempty"`
	// Use rke2-agent for RKE2 workers, or rke2-server for RKE2 servers.
	ServiceName string `json:"service_name,omitempty"`
	// Optional command prefix, e.g. nsenter --target 1 --mount -- systemctl.
	// Both activeness and uptime must query the same service manager.
	SystemctlCommands string `json:"systemctl_commands,omitempty"`
}

type ToolOverwrites struct {
	InfinibandClassRootDir string `json:"infiniband_class_root_dir"`

	// ExcludedInfinibandDevices is a list of InfiniBand device names to exclude from monitoring.
	// Device names should be like "mlx5_0", "mlx5_1", etc. (not full paths).
	//
	// This is useful for excluding devices that have restricted Physical Functions (PFs)
	// and cause kernel errors (mlx5_cmd_out_err ACCESS_REG) when queried.
	// This is common on NVIDIA DGX, Umbriel, and GB200 systems with ConnectX-7 adapters
	// where some ports are managed by system firmware.
	//
	// Example: ["mlx5_0", "mlx5_1"]
	//
	// ref.
	// https://github.com/prometheus/node_exporter/issues/3434
	// https://github.com/leptonai/gpud/issues/1164
	ExcludedInfinibandDevices []string `json:"excluded_infiniband_devices"`
}