Documentation
¶
Index ¶
- Variables
- func InitMetrics(config types.Config, rmgr resource.Manager, ...) error
- type Metrics
- func (m *Metrics) RemoveInvalidNodes(invalidNode string)
- func (m *Metrics) ResourceMiddleware(ctx context.Context, cluster cluster.Cluster) func(http.Handler) http.Handler
- func (m *Metrics) SendDeployCount(ctx context.Context, n int)
- func (m *Metrics) SendMetrics(ctx context.Context, metrics ...*plugintypes.Metrics)
- func (m *Metrics) SendNodesMetrics(ctx context.Context, nodes ...*types.Node)
- func (m *Metrics) SendPodNodeStatus(ctx context.Context, node *types.Node)
Constants ¶
This section is empty.
Variables ¶
View Source
var (
Client = Metrics{}
)
Functions ¶
func InitMetrics ¶
func InitMetrics(config types.Config, rmgr resource.Manager, metricsDescriptions []*plugintypes.MetricsDescription) error
InitMetrics builds the global metrics client and registers its collectors.
Types ¶
type Metrics ¶
type Metrics struct {
StatsdAddr string
Hostname string
Collectors map[string]prometheus.Collector
// contains filtered or unexported fields
}
Metrics ships core metrics to Prometheus and statsd.
func (*Metrics) RemoveInvalidNodes ¶
RemoveInvalidNodes drops Prometheus label sets for a node that no longer exists.
func (*Metrics) ResourceMiddleware ¶
func (m *Metrics) ResourceMiddleware(ctx context.Context, cluster cluster.Cluster) func(http.Handler) http.Handler
ResourceMiddleware refreshes node metrics before the wrapped handler runs; overlapping scrapes share one refresh, which runs under the server's context.
func (*Metrics) SendMetrics ¶
func (m *Metrics) SendMetrics(ctx context.Context, metrics ...*plugintypes.Metrics)
func (*Metrics) SendNodesMetrics ¶ added in v0.1.4
SendNodesMetrics refreshes the resource metrics of the nodes in one plugin call and pushes them.
Source Files
Click to show internal directories.
Click to hide internal directories.