API Server View

The API server view displays information about requests, work queues, and resources, so you can better monitor the API server.

API server metrics include request metrics, work queue metrics, and resource metrics. You can view these metrics in the following tables.

Figure 1 Request metrics
Click to enlarge

**Table 1** Request metrics
Metric	Unit	Description
Alived	N/A	The number of live component instances.
QPS	request/s	The number of requests with different response codes per second.
Request Success Rate (Read)	%	The percentage of read requests whose response code is 20x per second.
Requests Being Processed	N/A	The number of requests being processed by the API server.
Request Rate (Read)	request/s	The number of read requests with different response codes per second.
Request Error Rate (Read)	%	The percentage of read requests that are rejected per second.
P99 Request Latency (Read)	ms	P99 latency of read operations.
Request Rate (Write)	request/s	The number of write requests with different response codes per second.
Request Error Rate (Write)	%	The percentage of write requests that are rejected per second.
P99 Request Latency (Write)	ms	P99 latency of write operations.

Figure 2 Work queue metrics
Click to enlarge

**Table 2** Work queue metrics
Metric	Unit	Description
Work Queue Growth Rate	N/A	The number of additions handled by the API server work queue per second.
Work Queue Depth	N/A	Depth of a work queue.
Work Queue Latency (P99)	ms	P99 latency of each API server request in a work queue.

Figure 3 Resource metrics
Click to enlarge

The following is the metric list of the API server view.

**Table 4** Metric description
Metric	Type	Description
up	Gauge	Component status.
apiserver_request_total	Counter	Counter of API server requests broken out for code and other items.
go_goroutines	Gauge	The number of goroutines that currently exist.
apiserver_current_inflight_requests	Gauge	Maximum number of requests that were being actively served in the last one-second window.
apiserver_request_duration_seconds_bucket	Histogram	Latency for each request to the API server in seconds.
workqueue_depth	Gauge	Current depth of a work queue.
workqueue_adds_total	Counter	Total number of additions handled by a work queue.
workqueue_queue_duration_seconds_bucket	Histogram	How long in seconds an item stays in a work queue before being requested.
process_resident_memory_bytes	Gauge	Resident memory size in bytes.
process_cpu_seconds_total	Counter	Total user and system CPU time spent in seconds.