Documentation
¶
Index ¶
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type CgroupCPUWatcher ¶
type CgroupCPUWatcher struct {
// contains filtered or unexported fields
}
CgroupCPUWatcher implements ResourceWatcher interface for watching CPU throttling of cgroup. Cgroup doesn't have an easy way to gauge how bad the CPU is percentage. Hence, this watcher compares the recorded total throttled time between two polls. If the throttled time exceeds 50% of the observation window, it returns a backoff event. The watcher uses `throttled_time` (CgroupV1) or `throttled_usec` (CgroupV2) stats from the cgroup manager.
func NewCgroupCPUWatcher ¶
func NewCgroupCPUWatcher(manager cgroups.Manager, cpuThrottledThreshold float64) *CgroupCPUWatcher
NewCgroupCPUWatcher is the initializer of CgroupCPUWatcher
func (*CgroupCPUWatcher) Name ¶
func (c *CgroupCPUWatcher) Name() string
Name returns the name of CgroupCPUWatcher
func (*CgroupCPUWatcher) Poll ¶
func (c *CgroupCPUWatcher) Poll(ctx context.Context) (*limiter.BackoffEvent, error)
Poll asserts the cgroup statistics and returns a backoff event accordingly. The condition when a backoff event is returned is described above.
type CgroupMemoryWatcher ¶
type CgroupMemoryWatcher struct {
// contains filtered or unexported fields
}
CgroupMemoryWatcher implements ResourceWatcher interface. This watcher polls the statistics from the cgroup manager. It returns a backoff event in two conditions: * The current memory usage exceeds a soft threshold (90%). * The cgroup is under OOM.
func NewCgroupMemoryWatcher ¶
func NewCgroupMemoryWatcher(manager cgroups.Manager, memoryThreshold float64) *CgroupMemoryWatcher
NewCgroupMemoryWatcher is the initializer of CgroupMemoryWatcher
func (*CgroupMemoryWatcher) Name ¶
func (c *CgroupMemoryWatcher) Name() string
Name returns the name of CgroupMemoryWatcher
func (*CgroupMemoryWatcher) Poll ¶
func (c *CgroupMemoryWatcher) Poll(context.Context) (*limiter.BackoffEvent, error)
Poll asserts the cgroup statistics and returns a backoff event accordingly when it is triggered. These stats are fetched from cgroup manager.
Source Files