Documentation ¶
Overview ¶
Package wait provides tools for polling or listening for changes to a condition.
Index ¶
- Variables
- func BackoffUntil(f func(), backoff BackoffManager, sliding bool, stopCh <-chan struct{})
- func ExponentialBackoff(backoff Backoff, condition ConditionFunc) error
- func Forever(f func(), period time.Duration)
- func Jitter(duration time.Duration, maxFactor float64) time.Duration
- func JitterUntil(f func(), period time.Duration, jitterFactor float64, sliding bool, ...)
- func JitterUntilWithContext(ctx context.Context, f func(context.Context), period time.Duration, ...)
- func NonSlidingUntil(f func(), period time.Duration, stopCh <-chan struct{})
- func NonSlidingUntilWithContext(ctx context.Context, f func(context.Context), period time.Duration)
- func Poll(interval, timeout time.Duration, condition ConditionFunc) error
- func PollImmediate(interval, timeout time.Duration, condition ConditionFunc) error
- func PollImmediateInfinite(interval time.Duration, condition ConditionFunc) error
- func PollImmediateUntil(interval time.Duration, condition ConditionFunc, stopCh <-chan struct{}) error
- func PollInfinite(interval time.Duration, condition ConditionFunc) error
- func PollUntil(interval time.Duration, condition ConditionFunc, stopCh <-chan struct{}) error
- func Until(f func(), period time.Duration, stopCh <-chan struct{})
- func UntilWithContext(ctx context.Context, f func(context.Context), period time.Duration)
- func WaitFor(wait WaitFunc, fn ConditionFunc, done <-chan struct{}) error
- type Backoff
- type BackoffManager
- type ConditionFunc
- type Group
- type WaitFunc
Constants ¶
This section is empty.
Variables ¶
ErrWaitTimeout is returned when the condition exited without success.
For any test of the style:
... <- time.After(timeout): t.Errorf("Timed out")
The value for timeout should effectively be "forever." Obviously we don't want our tests to truly lock up forever, but 30s is long enough that it is effectively forever for the things that can slow down a run on a heavily contended machine (GC, seeks, etc), but not so long as to make a developer ctrl-c a test run if they do happen to break that test.
var NeverStop <-chan struct{} = make(chan struct{})
NeverStop may be passed to Until to make it never stop.
Functions ¶
func BackoffUntil ¶
func BackoffUntil(f func(), backoff BackoffManager, sliding bool, stopCh <-chan struct{})
BackoffUntil loops until stop channel is closed, run f every duration given by BackoffManager.
If sliding is true, the period is computed after f runs. If it is false then period includes the runtime for f.
func ExponentialBackoff ¶
func ExponentialBackoff(backoff Backoff, condition ConditionFunc) error
ExponentialBackoff repeats a condition check with exponential backoff.
It repeatedly checks the condition and then sleeps, using `backoff.Step()` to determine the length of the sleep and adjust Duration and Steps. Stops and returns as soon as: 1. the condition check returns true or an error, 2. `backoff.Steps` checks of the condition have been done, or 3. a sleep truncated by the cap on duration has been completed. In case (1) the returned error is what the condition function returned. In all other cases, ErrWaitTimeout is returned.
func Jitter ¶
Jitter returns a time.Duration between duration and duration + maxFactor * duration.
This allows clients to avoid converging on periodic behavior. If maxFactor is 0.0, a suggested default value will be chosen.
func JitterUntil ¶
func JitterUntil(f func(), period time.Duration, jitterFactor float64, sliding bool, stopCh <-chan struct{})
JitterUntil loops until stop channel is closed, running f every period.
If jitterFactor is positive, the period is jittered before every run of f. If jitterFactor is not positive, the period is unchanged and not jittered.
If sliding is true, the period is computed after f runs. If it is false then period includes the runtime for f.
Close stopCh to stop. f may not be invoked if stop channel is already closed. Pass NeverStop to if you don't want it stop.
func JitterUntilWithContext ¶
func JitterUntilWithContext(ctx context.Context, f func(context.Context), period time.Duration, jitterFactor float64, sliding bool)
JitterUntilWithContext loops until context is done, running f every period.
If jitterFactor is positive, the period is jittered before every run of f. If jitterFactor is not positive, the period is unchanged and not jittered.
If sliding is true, the period is computed after f runs. If it is false then period includes the runtime for f.
Cancel context to stop. f may not be invoked if context is already expired.
func NonSlidingUntil ¶
NonSlidingUntil loops until stop channel is closed, running f every period.
NonSlidingUntil is syntactic sugar on top of JitterUntil with zero jitter factor, with sliding = false (meaning the timer for period starts at the same time as the function starts).
func NonSlidingUntilWithContext ¶
NonSlidingUntilWithContext loops until context is done, running f every period.
NonSlidingUntilWithContext is syntactic sugar on top of JitterUntilWithContext with zero jitter factor, with sliding = false (meaning the timer for period starts at the same time as the function starts).
func Poll ¶
func Poll(interval, timeout time.Duration, condition ConditionFunc) error
Poll tries a condition func until it returns true, an error, or the timeout is reached.
Poll always waits the interval before the run of 'condition'. 'condition' will always be invoked at least once.
Some intervals may be missed if the condition takes too long or the time window is too short.
If you want to Poll something forever, see PollInfinite.
func PollImmediate ¶
func PollImmediate(interval, timeout time.Duration, condition ConditionFunc) error
PollImmediate tries a condition func until it returns true, an error, or the timeout is reached.
PollImmediate always checks 'condition' before waiting for the interval. 'condition' will always be invoked at least once.
Some intervals may be missed if the condition takes too long or the time window is too short.
If you want to immediately Poll something forever, see PollImmediateInfinite.
func PollImmediateInfinite ¶
func PollImmediateInfinite(interval time.Duration, condition ConditionFunc) error
PollImmediateInfinite tries a condition func until it returns true or an error
PollImmediateInfinite runs the 'condition' before waiting for the interval.
Some intervals may be missed if the condition takes too long or the time window is too short.
func PollImmediateUntil ¶
func PollImmediateUntil(interval time.Duration, condition ConditionFunc, stopCh <-chan struct{}) error
PollImmediateUntil tries a condition func until it returns true, an error or stopCh is closed.
PollImmediateUntil runs the 'condition' before waiting for the interval. 'condition' will always be invoked at least once.
func PollInfinite ¶
func PollInfinite(interval time.Duration, condition ConditionFunc) error
PollInfinite tries a condition func until it returns true or an error
PollInfinite always waits the interval before the run of 'condition'.
Some intervals may be missed if the condition takes too long or the time window is too short.
func PollUntil ¶
func PollUntil(interval time.Duration, condition ConditionFunc, stopCh <-chan struct{}) error
PollUntil tries a condition func until it returns true, an error or stopCh is closed.
PollUntil always waits interval before the first run of 'condition'. 'condition' will always be invoked at least once.
func Until ¶
Until loops until stop channel is closed, running f every period.
Until is syntactic sugar on top of JitterUntil with zero jitter factor and with sliding = true (which means the timer for period starts after the f completes).
func UntilWithContext ¶
UntilWithContext loops until context is done, running f every period.
UntilWithContext is syntactic sugar on top of JitterUntilWithContext with zero jitter factor and with sliding = true (which means the timer for period starts after the f completes).
func WaitFor ¶
func WaitFor(wait WaitFunc, fn ConditionFunc, done <-chan struct{}) error
WaitFor continually checks 'fn' as driven by 'wait'.
WaitFor gets a channel from 'wait()”, and then invokes 'fn' once for every value placed on the channel and once more when the channel is closed. If the channel is closed and 'fn' returns false without error, WaitFor returns ErrWaitTimeout.
If 'fn' returns an error the loop ends and that error is returned. If 'fn' returns true the loop ends and nil is returned.
ErrWaitTimeout will be returned if the 'done' channel is closed without fn ever returning true.
When the done channel is closed, because the golang `select` statement is "uniform pseudo-random", the `fn` might still run one or multiple time, though eventually `WaitFor` will return.
Types ¶
type Backoff ¶
type Backoff struct { // The initial duration. Duration time.Duration // Duration is multiplied by factor each iteration, if factor is not zero // and the limits imposed by Steps and Cap have not been reached. // Should not be negative. // The jitter does not contribute to the updates to the duration parameter. Factor float64 // The sleep at each iteration is the duration plus an additional // amount chosen uniformly at random from the interval between // zero and `jitter*duration`. Jitter float64 // The remaining number of iterations in which the duration // parameter may change (but progress can be stopped earlier by // hitting the cap). If not positive, the duration is not // changed. Used for exponential backoff in combination with // Factor and Cap. Steps int // A limit on revised values of the duration parameter. If a // multiplication by the factor parameter would make the duration // exceed the cap then the duration is set to the cap and the // steps parameter is set to zero. Cap time.Duration }
Backoff holds parameters applied to a Backoff function.
type BackoffManager ¶
BackoffManager manages backoff with a particular scheme based on its underlying implementation. It provides an interface to return a timer for backoff, and caller shall backoff until Timer.C() drains. If the second Backoff() is called before the timer from the first Backoff() call finishes, the first timer will NOT be drained and result in undetermined behavior. The BackoffManager is supposed to be called in a single-threaded environment.
func NewExponentialBackoffManager ¶
func NewExponentialBackoffManager(initBackoff, maxBackoff, resetDuration time.Duration, backoffFactor, jitter float64, c clock.Clock) BackoffManager
NewExponentialBackoffManager returns a manager for managing exponential backoff. Each backoff is jittered and backoff will not exceed the given max. If the backoff is not called within resetDuration, the backoff is reset. This backoff manager is used to reduce load during upstream unhealthiness.
type ConditionFunc ¶
ConditionFunc returns true if the condition is satisfied, or an error if the loop should be aborted.
type Group ¶
type Group struct {
// contains filtered or unexported fields
}
Group allows to start a group of goroutines and wait for their completion.
func (*Group) Start ¶
func (g *Group) Start(f func())
Start starts f in a new goroutine in the group.
func (*Group) StartWithChannel ¶
func (g *Group) StartWithChannel(stopCh <-chan struct{}, f func(stopCh <-chan struct{}))
StartWithChannel starts f in a new goroutine in the group. stopCh is passed to f as an argument. f should stop when stopCh is available.
func (*Group) StartWithContext ¶
StartWithContext starts f in a new goroutine in the group. ctx is passed to f as an argument. f should stop when ctx.Done() is available.
type WaitFunc ¶
type WaitFunc func(done <-chan struct{}) <-chan struct{}
WaitFunc creates a channel that receives an item every time a test should be executed and is closed when the last test should be invoked.