mirror of
https://github.com/go-gitea/gitea
synced 2024-10-31 23:34:25 +00:00
Multiple Queue improvements: LevelDB Wait on empty, shutdown empty shadow level queue, reduce goroutines etc (#15693)
* move shutdownfns, terminatefns and hammerfns out of separate goroutines Coalesce the shutdownfns etc into a list of functions that get run at shutdown rather then have them run at goroutines blocked on selects. This may help reduce the background select/poll load in certain configurations. * The LevelDB queues can actually wait on empty instead of polling Slight refactor to cause leveldb queues to wait on empty instead of polling. * Shutdown the shadow level queue once it is empty * Remove bytefifo additional goroutine for readToChan as it can just be run in run * Remove additional removeWorkers goroutine for workers * Simplify the AtShutdown and AtTerminate functions and add Channel Flusher * Add shutdown flusher to CUQ * move persistable channel shutdown stuff to Shutdown Fn * Ensure that UPCQ has the correct config * handle shutdown during the flushing * reduce risk of race between zeroBoost and addWorkers * prevent double shutdown Signed-off-by: Andrew Thornton <art27@cantab.net>
This commit is contained in:
parent
9f19c2b8cc
commit
ba526ceffe
@ -6,17 +6,9 @@ package graceful
|
|||||||
|
|
||||||
import (
|
import (
|
||||||
"context"
|
"context"
|
||||||
"fmt"
|
|
||||||
"time"
|
"time"
|
||||||
)
|
)
|
||||||
|
|
||||||
// Errors for context.Err()
|
|
||||||
var (
|
|
||||||
ErrShutdown = fmt.Errorf("Graceful Manager called Shutdown")
|
|
||||||
ErrHammer = fmt.Errorf("Graceful Manager called Hammer")
|
|
||||||
ErrTerminate = fmt.Errorf("Graceful Manager called Terminate")
|
|
||||||
)
|
|
||||||
|
|
||||||
// ChannelContext is a context that wraps a channel and error as a context
|
// ChannelContext is a context that wraps a channel and error as a context
|
||||||
type ChannelContext struct {
|
type ChannelContext struct {
|
||||||
done <-chan struct{}
|
done <-chan struct{}
|
||||||
@ -63,28 +55,19 @@ func (ctx *ChannelContext) Value(key interface{}) interface{} {
|
|||||||
// Callers using this context should ensure that they are registered as a running server
|
// Callers using this context should ensure that they are registered as a running server
|
||||||
// in order that they are waited for.
|
// in order that they are waited for.
|
||||||
func (g *Manager) ShutdownContext() context.Context {
|
func (g *Manager) ShutdownContext() context.Context {
|
||||||
return &ChannelContext{
|
return g.shutdownCtx
|
||||||
done: g.IsShutdown(),
|
|
||||||
err: ErrShutdown,
|
|
||||||
}
|
|
||||||
}
|
}
|
||||||
|
|
||||||
// HammerContext returns a context.Context that is Done at hammer
|
// HammerContext returns a context.Context that is Done at hammer
|
||||||
// Callers using this context should ensure that they are registered as a running server
|
// Callers using this context should ensure that they are registered as a running server
|
||||||
// in order that they are waited for.
|
// in order that they are waited for.
|
||||||
func (g *Manager) HammerContext() context.Context {
|
func (g *Manager) HammerContext() context.Context {
|
||||||
return &ChannelContext{
|
return g.hammerCtx
|
||||||
done: g.IsHammer(),
|
|
||||||
err: ErrHammer,
|
|
||||||
}
|
|
||||||
}
|
}
|
||||||
|
|
||||||
// TerminateContext returns a context.Context that is Done at terminate
|
// TerminateContext returns a context.Context that is Done at terminate
|
||||||
// Callers using this context should ensure that they are registered as a terminating server
|
// Callers using this context should ensure that they are registered as a terminating server
|
||||||
// in order that they are waited for.
|
// in order that they are waited for.
|
||||||
func (g *Manager) TerminateContext() context.Context {
|
func (g *Manager) TerminateContext() context.Context {
|
||||||
return &ChannelContext{
|
return g.terminateCtx
|
||||||
done: g.IsTerminate(),
|
|
||||||
err: ErrTerminate,
|
|
||||||
}
|
|
||||||
}
|
}
|
||||||
|
@ -54,8 +54,8 @@ func InitManager(ctx context.Context) {
|
|||||||
})
|
})
|
||||||
}
|
}
|
||||||
|
|
||||||
// CallbackWithContext is combined runnable and context to watch to see if the caller has finished
|
// WithCallback is a runnable to call when the caller has finished
|
||||||
type CallbackWithContext func(ctx context.Context, callback func())
|
type WithCallback func(callback func())
|
||||||
|
|
||||||
// RunnableWithShutdownFns is a runnable with functions to run at shutdown and terminate
|
// RunnableWithShutdownFns is a runnable with functions to run at shutdown and terminate
|
||||||
// After the callback to atShutdown is called and is complete, the main function must return.
|
// After the callback to atShutdown is called and is complete, the main function must return.
|
||||||
@ -63,7 +63,7 @@ type CallbackWithContext func(ctx context.Context, callback func())
|
|||||||
// Please note that use of the atShutdown and atTerminate callbacks will create go-routines that will wait till their respective signals
|
// Please note that use of the atShutdown and atTerminate callbacks will create go-routines that will wait till their respective signals
|
||||||
// - users must therefore be careful to only call these as necessary.
|
// - users must therefore be careful to only call these as necessary.
|
||||||
// If run is not expected to run indefinitely RunWithShutdownChan is likely to be more appropriate.
|
// If run is not expected to run indefinitely RunWithShutdownChan is likely to be more appropriate.
|
||||||
type RunnableWithShutdownFns func(atShutdown, atTerminate func(context.Context, func()))
|
type RunnableWithShutdownFns func(atShutdown, atTerminate func(func()))
|
||||||
|
|
||||||
// RunWithShutdownFns takes a function that has both atShutdown and atTerminate callbacks
|
// RunWithShutdownFns takes a function that has both atShutdown and atTerminate callbacks
|
||||||
// After the callback to atShutdown is called and is complete, the main function must return.
|
// After the callback to atShutdown is called and is complete, the main function must return.
|
||||||
@ -80,17 +80,21 @@ func (g *Manager) RunWithShutdownFns(run RunnableWithShutdownFns) {
|
|||||||
g.doShutdown()
|
g.doShutdown()
|
||||||
}
|
}
|
||||||
}()
|
}()
|
||||||
run(func(ctx context.Context, atShutdown func()) {
|
run(func(atShutdown func()) {
|
||||||
go func() {
|
g.lock.Lock()
|
||||||
select {
|
defer g.lock.Unlock()
|
||||||
case <-g.IsShutdown():
|
g.toRunAtShutdown = append(g.toRunAtShutdown,
|
||||||
|
func() {
|
||||||
|
defer func() {
|
||||||
|
if err := recover(); err != nil {
|
||||||
|
log.Critical("PANIC during RunWithShutdownFns: %v\nStacktrace: %s", err, log.Stack(2))
|
||||||
|
g.doShutdown()
|
||||||
|
}
|
||||||
|
}()
|
||||||
atShutdown()
|
atShutdown()
|
||||||
case <-ctx.Done():
|
})
|
||||||
return
|
}, func(atTerminate func()) {
|
||||||
}
|
g.RunAtTerminate(atTerminate)
|
||||||
}()
|
|
||||||
}, func(ctx context.Context, atTerminate func()) {
|
|
||||||
g.RunAtTerminate(ctx, atTerminate)
|
|
||||||
})
|
})
|
||||||
}
|
}
|
||||||
|
|
||||||
@ -99,7 +103,7 @@ func (g *Manager) RunWithShutdownFns(run RunnableWithShutdownFns) {
|
|||||||
// (Optionally IsHammer may be waited for instead however, this should be avoided if possible.)
|
// (Optionally IsHammer may be waited for instead however, this should be avoided if possible.)
|
||||||
// The callback function provided to atTerminate must return once termination is complete.
|
// The callback function provided to atTerminate must return once termination is complete.
|
||||||
// Please note that use of the atTerminate function will create a go-routine that will wait till terminate - users must therefore be careful to only call this as necessary.
|
// Please note that use of the atTerminate function will create a go-routine that will wait till terminate - users must therefore be careful to only call this as necessary.
|
||||||
type RunnableWithShutdownChan func(atShutdown <-chan struct{}, atTerminate CallbackWithContext)
|
type RunnableWithShutdownChan func(atShutdown <-chan struct{}, atTerminate WithCallback)
|
||||||
|
|
||||||
// RunWithShutdownChan takes a function that has channel to watch for shutdown and atTerminate callbacks
|
// RunWithShutdownChan takes a function that has channel to watch for shutdown and atTerminate callbacks
|
||||||
// After the atShutdown channel is closed, the main function must return once shutdown is complete.
|
// After the atShutdown channel is closed, the main function must return once shutdown is complete.
|
||||||
@ -115,8 +119,8 @@ func (g *Manager) RunWithShutdownChan(run RunnableWithShutdownChan) {
|
|||||||
g.doShutdown()
|
g.doShutdown()
|
||||||
}
|
}
|
||||||
}()
|
}()
|
||||||
run(g.IsShutdown(), func(ctx context.Context, atTerminate func()) {
|
run(g.IsShutdown(), func(atTerminate func()) {
|
||||||
g.RunAtTerminate(ctx, atTerminate)
|
g.RunAtTerminate(atTerminate)
|
||||||
})
|
})
|
||||||
}
|
}
|
||||||
|
|
||||||
@ -136,60 +140,65 @@ func (g *Manager) RunWithShutdownContext(run func(context.Context)) {
|
|||||||
}
|
}
|
||||||
|
|
||||||
// RunAtTerminate adds to the terminate wait group and creates a go-routine to run the provided function at termination
|
// RunAtTerminate adds to the terminate wait group and creates a go-routine to run the provided function at termination
|
||||||
func (g *Manager) RunAtTerminate(ctx context.Context, terminate func()) {
|
func (g *Manager) RunAtTerminate(terminate func()) {
|
||||||
g.terminateWaitGroup.Add(1)
|
g.terminateWaitGroup.Add(1)
|
||||||
go func() {
|
g.lock.Lock()
|
||||||
defer g.terminateWaitGroup.Done()
|
defer g.lock.Unlock()
|
||||||
defer func() {
|
g.toRunAtTerminate = append(g.toRunAtTerminate,
|
||||||
if err := recover(); err != nil {
|
func() {
|
||||||
log.Critical("PANIC during RunAtTerminate: %v\nStacktrace: %s", err, log.Stack(2))
|
defer g.terminateWaitGroup.Done()
|
||||||
}
|
defer func() {
|
||||||
}()
|
if err := recover(); err != nil {
|
||||||
select {
|
log.Critical("PANIC during RunAtTerminate: %v\nStacktrace: %s", err, log.Stack(2))
|
||||||
case <-g.IsTerminate():
|
}
|
||||||
|
}()
|
||||||
terminate()
|
terminate()
|
||||||
case <-ctx.Done():
|
})
|
||||||
}
|
|
||||||
}()
|
|
||||||
}
|
}
|
||||||
|
|
||||||
// RunAtShutdown creates a go-routine to run the provided function at shutdown
|
// RunAtShutdown creates a go-routine to run the provided function at shutdown
|
||||||
func (g *Manager) RunAtShutdown(ctx context.Context, shutdown func()) {
|
func (g *Manager) RunAtShutdown(ctx context.Context, shutdown func()) {
|
||||||
go func() {
|
g.lock.Lock()
|
||||||
defer func() {
|
defer g.lock.Unlock()
|
||||||
if err := recover(); err != nil {
|
g.toRunAtShutdown = append(g.toRunAtShutdown,
|
||||||
log.Critical("PANIC during RunAtShutdown: %v\nStacktrace: %s", err, log.Stack(2))
|
func() {
|
||||||
|
defer func() {
|
||||||
|
if err := recover(); err != nil {
|
||||||
|
log.Critical("PANIC during RunAtShutdown: %v\nStacktrace: %s", err, log.Stack(2))
|
||||||
|
}
|
||||||
|
}()
|
||||||
|
select {
|
||||||
|
case <-ctx.Done():
|
||||||
|
return
|
||||||
|
default:
|
||||||
|
shutdown()
|
||||||
}
|
}
|
||||||
}()
|
})
|
||||||
select {
|
|
||||||
case <-g.IsShutdown():
|
|
||||||
shutdown()
|
|
||||||
case <-ctx.Done():
|
|
||||||
}
|
|
||||||
}()
|
|
||||||
}
|
}
|
||||||
|
|
||||||
// RunAtHammer creates a go-routine to run the provided function at shutdown
|
// RunAtHammer creates a go-routine to run the provided function at shutdown
|
||||||
func (g *Manager) RunAtHammer(ctx context.Context, hammer func()) {
|
func (g *Manager) RunAtHammer(hammer func()) {
|
||||||
go func() {
|
g.lock.Lock()
|
||||||
defer func() {
|
defer g.lock.Unlock()
|
||||||
if err := recover(); err != nil {
|
g.toRunAtHammer = append(g.toRunAtHammer,
|
||||||
log.Critical("PANIC during RunAtHammer: %v\nStacktrace: %s", err, log.Stack(2))
|
func() {
|
||||||
}
|
defer func() {
|
||||||
}()
|
if err := recover(); err != nil {
|
||||||
select {
|
log.Critical("PANIC during RunAtHammer: %v\nStacktrace: %s", err, log.Stack(2))
|
||||||
case <-g.IsHammer():
|
}
|
||||||
|
}()
|
||||||
hammer()
|
hammer()
|
||||||
case <-ctx.Done():
|
})
|
||||||
}
|
|
||||||
}()
|
|
||||||
}
|
}
|
||||||
func (g *Manager) doShutdown() {
|
func (g *Manager) doShutdown() {
|
||||||
if !g.setStateTransition(stateRunning, stateShuttingDown) {
|
if !g.setStateTransition(stateRunning, stateShuttingDown) {
|
||||||
return
|
return
|
||||||
}
|
}
|
||||||
g.lock.Lock()
|
g.lock.Lock()
|
||||||
close(g.shutdown)
|
g.shutdownCtxCancel()
|
||||||
|
for _, fn := range g.toRunAtShutdown {
|
||||||
|
go fn()
|
||||||
|
}
|
||||||
g.lock.Unlock()
|
g.lock.Unlock()
|
||||||
|
|
||||||
if setting.GracefulHammerTime >= 0 {
|
if setting.GracefulHammerTime >= 0 {
|
||||||
@ -203,7 +212,7 @@ func (g *Manager) doShutdown() {
|
|||||||
g.doTerminate()
|
g.doTerminate()
|
||||||
g.WaitForTerminate()
|
g.WaitForTerminate()
|
||||||
g.lock.Lock()
|
g.lock.Lock()
|
||||||
close(g.done)
|
g.doneCtxCancel()
|
||||||
g.lock.Unlock()
|
g.lock.Unlock()
|
||||||
}()
|
}()
|
||||||
}
|
}
|
||||||
@ -212,10 +221,13 @@ func (g *Manager) doHammerTime(d time.Duration) {
|
|||||||
time.Sleep(d)
|
time.Sleep(d)
|
||||||
g.lock.Lock()
|
g.lock.Lock()
|
||||||
select {
|
select {
|
||||||
case <-g.hammer:
|
case <-g.hammerCtx.Done():
|
||||||
default:
|
default:
|
||||||
log.Warn("Setting Hammer condition")
|
log.Warn("Setting Hammer condition")
|
||||||
close(g.hammer)
|
g.hammerCtxCancel()
|
||||||
|
for _, fn := range g.toRunAtHammer {
|
||||||
|
go fn()
|
||||||
|
}
|
||||||
}
|
}
|
||||||
g.lock.Unlock()
|
g.lock.Unlock()
|
||||||
}
|
}
|
||||||
@ -226,10 +238,13 @@ func (g *Manager) doTerminate() {
|
|||||||
}
|
}
|
||||||
g.lock.Lock()
|
g.lock.Lock()
|
||||||
select {
|
select {
|
||||||
case <-g.terminate:
|
case <-g.terminateCtx.Done():
|
||||||
default:
|
default:
|
||||||
log.Warn("Terminating")
|
log.Warn("Terminating")
|
||||||
close(g.terminate)
|
g.terminateCtxCancel()
|
||||||
|
for _, fn := range g.toRunAtTerminate {
|
||||||
|
go fn()
|
||||||
|
}
|
||||||
}
|
}
|
||||||
g.lock.Unlock()
|
g.lock.Unlock()
|
||||||
}
|
}
|
||||||
@ -242,7 +257,7 @@ func (g *Manager) IsChild() bool {
|
|||||||
// IsShutdown returns a channel which will be closed at shutdown.
|
// IsShutdown returns a channel which will be closed at shutdown.
|
||||||
// The order of closure is IsShutdown, IsHammer (potentially), IsTerminate
|
// The order of closure is IsShutdown, IsHammer (potentially), IsTerminate
|
||||||
func (g *Manager) IsShutdown() <-chan struct{} {
|
func (g *Manager) IsShutdown() <-chan struct{} {
|
||||||
return g.shutdown
|
return g.shutdownCtx.Done()
|
||||||
}
|
}
|
||||||
|
|
||||||
// IsHammer returns a channel which will be closed at hammer
|
// IsHammer returns a channel which will be closed at hammer
|
||||||
@ -250,14 +265,14 @@ func (g *Manager) IsShutdown() <-chan struct{} {
|
|||||||
// Servers running within the running server wait group should respond to IsHammer
|
// Servers running within the running server wait group should respond to IsHammer
|
||||||
// if not shutdown already
|
// if not shutdown already
|
||||||
func (g *Manager) IsHammer() <-chan struct{} {
|
func (g *Manager) IsHammer() <-chan struct{} {
|
||||||
return g.hammer
|
return g.hammerCtx.Done()
|
||||||
}
|
}
|
||||||
|
|
||||||
// IsTerminate returns a channel which will be closed at terminate
|
// IsTerminate returns a channel which will be closed at terminate
|
||||||
// The order of closure is IsShutdown, IsHammer (potentially), IsTerminate
|
// The order of closure is IsShutdown, IsHammer (potentially), IsTerminate
|
||||||
// IsTerminate will only close once all running servers have stopped
|
// IsTerminate will only close once all running servers have stopped
|
||||||
func (g *Manager) IsTerminate() <-chan struct{} {
|
func (g *Manager) IsTerminate() <-chan struct{} {
|
||||||
return g.terminate
|
return g.terminateCtx.Done()
|
||||||
}
|
}
|
||||||
|
|
||||||
// ServerDone declares a running server done and subtracts one from the
|
// ServerDone declares a running server done and subtracts one from the
|
||||||
@ -314,25 +329,20 @@ func (g *Manager) InformCleanup() {
|
|||||||
|
|
||||||
// Done allows the manager to be viewed as a context.Context, it returns a channel that is closed when the server is finished terminating
|
// Done allows the manager to be viewed as a context.Context, it returns a channel that is closed when the server is finished terminating
|
||||||
func (g *Manager) Done() <-chan struct{} {
|
func (g *Manager) Done() <-chan struct{} {
|
||||||
return g.done
|
return g.doneCtx.Done()
|
||||||
}
|
}
|
||||||
|
|
||||||
// Err allows the manager to be viewed as a context.Context done at Terminate, it returns ErrTerminate
|
// Err allows the manager to be viewed as a context.Context done at Terminate
|
||||||
func (g *Manager) Err() error {
|
func (g *Manager) Err() error {
|
||||||
select {
|
return g.doneCtx.Err()
|
||||||
case <-g.Done():
|
|
||||||
return ErrTerminate
|
|
||||||
default:
|
|
||||||
return nil
|
|
||||||
}
|
|
||||||
}
|
}
|
||||||
|
|
||||||
// Value allows the manager to be viewed as a context.Context done at Terminate, it has no values
|
// Value allows the manager to be viewed as a context.Context done at Terminate
|
||||||
func (g *Manager) Value(key interface{}) interface{} {
|
func (g *Manager) Value(key interface{}) interface{} {
|
||||||
return nil
|
return g.doneCtx.Value(key)
|
||||||
}
|
}
|
||||||
|
|
||||||
// Deadline returns nil as there is no fixed Deadline for the manager, it allows the manager to be viewed as a context.Context
|
// Deadline returns nil as there is no fixed Deadline for the manager, it allows the manager to be viewed as a context.Context
|
||||||
func (g *Manager) Deadline() (deadline time.Time, ok bool) {
|
func (g *Manager) Deadline() (deadline time.Time, ok bool) {
|
||||||
return
|
return g.doneCtx.Deadline()
|
||||||
}
|
}
|
||||||
|
@ -25,13 +25,21 @@ type Manager struct {
|
|||||||
forked bool
|
forked bool
|
||||||
lock *sync.RWMutex
|
lock *sync.RWMutex
|
||||||
state state
|
state state
|
||||||
shutdown chan struct{}
|
shutdownCtx context.Context
|
||||||
hammer chan struct{}
|
hammerCtx context.Context
|
||||||
terminate chan struct{}
|
terminateCtx context.Context
|
||||||
done chan struct{}
|
doneCtx context.Context
|
||||||
|
shutdownCtxCancel context.CancelFunc
|
||||||
|
hammerCtxCancel context.CancelFunc
|
||||||
|
terminateCtxCancel context.CancelFunc
|
||||||
|
doneCtxCancel context.CancelFunc
|
||||||
runningServerWaitGroup sync.WaitGroup
|
runningServerWaitGroup sync.WaitGroup
|
||||||
createServerWaitGroup sync.WaitGroup
|
createServerWaitGroup sync.WaitGroup
|
||||||
terminateWaitGroup sync.WaitGroup
|
terminateWaitGroup sync.WaitGroup
|
||||||
|
|
||||||
|
toRunAtShutdown []func()
|
||||||
|
toRunAtHammer []func()
|
||||||
|
toRunAtTerminate []func()
|
||||||
}
|
}
|
||||||
|
|
||||||
func newGracefulManager(ctx context.Context) *Manager {
|
func newGracefulManager(ctx context.Context) *Manager {
|
||||||
@ -45,11 +53,11 @@ func newGracefulManager(ctx context.Context) *Manager {
|
|||||||
}
|
}
|
||||||
|
|
||||||
func (g *Manager) start(ctx context.Context) {
|
func (g *Manager) start(ctx context.Context) {
|
||||||
// Make channels
|
// Make contexts
|
||||||
g.terminate = make(chan struct{})
|
g.terminateCtx, g.terminateCtxCancel = context.WithCancel(ctx)
|
||||||
g.shutdown = make(chan struct{})
|
g.shutdownCtx, g.shutdownCtxCancel = context.WithCancel(ctx)
|
||||||
g.hammer = make(chan struct{})
|
g.hammerCtx, g.hammerCtxCancel = context.WithCancel(ctx)
|
||||||
g.done = make(chan struct{})
|
g.doneCtx, g.doneCtxCancel = context.WithCancel(ctx)
|
||||||
|
|
||||||
// Set the running state & handle signals
|
// Set the running state & handle signals
|
||||||
g.setState(stateRunning)
|
g.setState(stateRunning)
|
||||||
|
@ -36,14 +36,22 @@ type Manager struct {
|
|||||||
isChild bool
|
isChild bool
|
||||||
lock *sync.RWMutex
|
lock *sync.RWMutex
|
||||||
state state
|
state state
|
||||||
shutdown chan struct{}
|
shutdownCtx context.Context
|
||||||
hammer chan struct{}
|
hammerCtx context.Context
|
||||||
terminate chan struct{}
|
terminateCtx context.Context
|
||||||
done chan struct{}
|
doneCtx context.Context
|
||||||
|
shutdownCtxCancel context.CancelFunc
|
||||||
|
hammerCtxCancel context.CancelFunc
|
||||||
|
terminateCtxCancel context.CancelFunc
|
||||||
|
doneCtxCancel context.CancelFunc
|
||||||
runningServerWaitGroup sync.WaitGroup
|
runningServerWaitGroup sync.WaitGroup
|
||||||
createServerWaitGroup sync.WaitGroup
|
createServerWaitGroup sync.WaitGroup
|
||||||
terminateWaitGroup sync.WaitGroup
|
terminateWaitGroup sync.WaitGroup
|
||||||
shutdownRequested chan struct{}
|
shutdownRequested chan struct{}
|
||||||
|
|
||||||
|
toRunAtShutdown []func()
|
||||||
|
toRunAtHammer []func()
|
||||||
|
toRunAtTerminate []func()
|
||||||
}
|
}
|
||||||
|
|
||||||
func newGracefulManager(ctx context.Context) *Manager {
|
func newGracefulManager(ctx context.Context) *Manager {
|
||||||
@ -58,11 +66,13 @@ func newGracefulManager(ctx context.Context) *Manager {
|
|||||||
}
|
}
|
||||||
|
|
||||||
func (g *Manager) start() {
|
func (g *Manager) start() {
|
||||||
|
// Make contexts
|
||||||
|
g.terminateCtx, g.terminateCtxCancel = context.WithCancel(g.ctx)
|
||||||
|
g.shutdownCtx, g.shutdownCtxCancel = context.WithCancel(g.ctx)
|
||||||
|
g.hammerCtx, g.hammerCtxCancel = context.WithCancel(g.ctx)
|
||||||
|
g.doneCtx, g.doneCtxCancel = context.WithCancel(g.ctx)
|
||||||
|
|
||||||
// Make channels
|
// Make channels
|
||||||
g.terminate = make(chan struct{})
|
|
||||||
g.shutdown = make(chan struct{})
|
|
||||||
g.hammer = make(chan struct{})
|
|
||||||
g.done = make(chan struct{})
|
|
||||||
g.shutdownRequested = make(chan struct{})
|
g.shutdownRequested = make(chan struct{})
|
||||||
|
|
||||||
// Set the running state
|
// Set the running state
|
||||||
@ -171,7 +181,7 @@ hammerLoop:
|
|||||||
default:
|
default:
|
||||||
log.Debug("Unexpected control request: %v", change.Cmd)
|
log.Debug("Unexpected control request: %v", change.Cmd)
|
||||||
}
|
}
|
||||||
case <-g.hammer:
|
case <-g.hammerCtx.Done():
|
||||||
break hammerLoop
|
break hammerLoop
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
@ -115,7 +115,13 @@ func Init() {
|
|||||||
|
|
||||||
ctx, cancel := context.WithCancel(context.Background())
|
ctx, cancel := context.WithCancel(context.Background())
|
||||||
|
|
||||||
graceful.GetManager().RunAtTerminate(ctx, func() {
|
graceful.GetManager().RunAtTerminate(func() {
|
||||||
|
select {
|
||||||
|
case <-ctx.Done():
|
||||||
|
return
|
||||||
|
default:
|
||||||
|
}
|
||||||
|
cancel()
|
||||||
log.Debug("Closing repository indexer")
|
log.Debug("Closing repository indexer")
|
||||||
indexer.Close()
|
indexer.Close()
|
||||||
log.Info("PID: %d Repository Indexer closed", os.Getpid())
|
log.Info("PID: %d Repository Indexer closed", os.Getpid())
|
||||||
|
@ -160,7 +160,7 @@ func InitIssueIndexer(syncReindex bool) {
|
|||||||
}
|
}
|
||||||
populate = !exist
|
populate = !exist
|
||||||
holder.set(issueIndexer)
|
holder.set(issueIndexer)
|
||||||
graceful.GetManager().RunAtTerminate(context.Background(), func() {
|
graceful.GetManager().RunAtTerminate(func() {
|
||||||
log.Debug("Closing issue indexer")
|
log.Debug("Closing issue indexer")
|
||||||
issueIndexer := holder.get()
|
issueIndexer := holder.get()
|
||||||
if issueIndexer != nil {
|
if issueIndexer != nil {
|
||||||
@ -170,7 +170,7 @@ func InitIssueIndexer(syncReindex bool) {
|
|||||||
})
|
})
|
||||||
log.Debug("Created Bleve Indexer")
|
log.Debug("Created Bleve Indexer")
|
||||||
case "elasticsearch":
|
case "elasticsearch":
|
||||||
graceful.GetManager().RunWithShutdownFns(func(_, atTerminate func(context.Context, func())) {
|
graceful.GetManager().RunWithShutdownFns(func(_, atTerminate func(func())) {
|
||||||
issueIndexer, err := NewElasticSearchIndexer(setting.Indexer.IssueConnStr, setting.Indexer.IssueIndexerName)
|
issueIndexer, err := NewElasticSearchIndexer(setting.Indexer.IssueConnStr, setting.Indexer.IssueIndexerName)
|
||||||
if err != nil {
|
if err != nil {
|
||||||
log.Fatal("Unable to initialize Elastic Search Issue Indexer at connection: %s Error: %v", setting.Indexer.IssueConnStr, err)
|
log.Fatal("Unable to initialize Elastic Search Issue Indexer at connection: %s Error: %v", setting.Indexer.IssueConnStr, err)
|
||||||
|
@ -4,14 +4,16 @@
|
|||||||
|
|
||||||
package queue
|
package queue
|
||||||
|
|
||||||
|
import "context"
|
||||||
|
|
||||||
// ByteFIFO defines a FIFO that takes a byte array
|
// ByteFIFO defines a FIFO that takes a byte array
|
||||||
type ByteFIFO interface {
|
type ByteFIFO interface {
|
||||||
// Len returns the length of the fifo
|
// Len returns the length of the fifo
|
||||||
Len() int64
|
Len(ctx context.Context) int64
|
||||||
// PushFunc pushes data to the end of the fifo and calls the callback if it is added
|
// PushFunc pushes data to the end of the fifo and calls the callback if it is added
|
||||||
PushFunc(data []byte, fn func() error) error
|
PushFunc(ctx context.Context, data []byte, fn func() error) error
|
||||||
// Pop pops data from the start of the fifo
|
// Pop pops data from the start of the fifo
|
||||||
Pop() ([]byte, error)
|
Pop(ctx context.Context) ([]byte, error)
|
||||||
// Close this fifo
|
// Close this fifo
|
||||||
Close() error
|
Close() error
|
||||||
}
|
}
|
||||||
@ -20,7 +22,7 @@ type ByteFIFO interface {
|
|||||||
type UniqueByteFIFO interface {
|
type UniqueByteFIFO interface {
|
||||||
ByteFIFO
|
ByteFIFO
|
||||||
// Has returns whether the fifo contains this data
|
// Has returns whether the fifo contains this data
|
||||||
Has(data []byte) (bool, error)
|
Has(ctx context.Context, data []byte) (bool, error)
|
||||||
}
|
}
|
||||||
|
|
||||||
var _ ByteFIFO = &DummyByteFIFO{}
|
var _ ByteFIFO = &DummyByteFIFO{}
|
||||||
@ -29,12 +31,12 @@ var _ ByteFIFO = &DummyByteFIFO{}
|
|||||||
type DummyByteFIFO struct{}
|
type DummyByteFIFO struct{}
|
||||||
|
|
||||||
// PushFunc returns nil
|
// PushFunc returns nil
|
||||||
func (*DummyByteFIFO) PushFunc(data []byte, fn func() error) error {
|
func (*DummyByteFIFO) PushFunc(ctx context.Context, data []byte, fn func() error) error {
|
||||||
return nil
|
return nil
|
||||||
}
|
}
|
||||||
|
|
||||||
// Pop returns nil
|
// Pop returns nil
|
||||||
func (*DummyByteFIFO) Pop() ([]byte, error) {
|
func (*DummyByteFIFO) Pop(ctx context.Context) ([]byte, error) {
|
||||||
return []byte{}, nil
|
return []byte{}, nil
|
||||||
}
|
}
|
||||||
|
|
||||||
@ -44,7 +46,7 @@ func (*DummyByteFIFO) Close() error {
|
|||||||
}
|
}
|
||||||
|
|
||||||
// Len is always 0
|
// Len is always 0
|
||||||
func (*DummyByteFIFO) Len() int64 {
|
func (*DummyByteFIFO) Len(ctx context.Context) int64 {
|
||||||
return 0
|
return 0
|
||||||
}
|
}
|
||||||
|
|
||||||
@ -56,6 +58,6 @@ type DummyUniqueByteFIFO struct {
|
|||||||
}
|
}
|
||||||
|
|
||||||
// Has always returns false
|
// Has always returns false
|
||||||
func (*DummyUniqueByteFIFO) Has([]byte) (bool, error) {
|
func (*DummyUniqueByteFIFO) Has(ctx context.Context, data []byte) (bool, error) {
|
||||||
return false, nil
|
return false, nil
|
||||||
}
|
}
|
||||||
|
@ -187,14 +187,14 @@ func (m *Manager) FlushAll(baseCtx context.Context, timeout time.Duration) error
|
|||||||
if flushable, ok := mq.Managed.(Flushable); ok {
|
if flushable, ok := mq.Managed.(Flushable); ok {
|
||||||
log.Debug("Flushing (flushable) queue: %s", mq.Name)
|
log.Debug("Flushing (flushable) queue: %s", mq.Name)
|
||||||
go func(q *ManagedQueue) {
|
go func(q *ManagedQueue) {
|
||||||
localCtx, localCancel := context.WithCancel(ctx)
|
localCtx, localCtxCancel := context.WithCancel(ctx)
|
||||||
pid := q.RegisterWorkers(1, start, hasTimeout, end, localCancel, true)
|
pid := q.RegisterWorkers(1, start, hasTimeout, end, localCtxCancel, true)
|
||||||
err := flushable.FlushWithContext(localCtx)
|
err := flushable.FlushWithContext(localCtx)
|
||||||
if err != nil && err != ctx.Err() {
|
if err != nil && err != ctx.Err() {
|
||||||
cancel()
|
cancel()
|
||||||
}
|
}
|
||||||
q.CancelWorkers(pid)
|
q.CancelWorkers(pid)
|
||||||
localCancel()
|
localCtxCancel()
|
||||||
wg.Done()
|
wg.Done()
|
||||||
}(mq)
|
}(mq)
|
||||||
} else {
|
} else {
|
||||||
|
@ -57,7 +57,7 @@ type Named interface {
|
|||||||
// Queues will handle their own contents in the Run method
|
// Queues will handle their own contents in the Run method
|
||||||
type Queue interface {
|
type Queue interface {
|
||||||
Flushable
|
Flushable
|
||||||
Run(atShutdown, atTerminate func(context.Context, func()))
|
Run(atShutdown, atTerminate func(func()))
|
||||||
Push(Data) error
|
Push(Data) error
|
||||||
}
|
}
|
||||||
|
|
||||||
@ -74,7 +74,7 @@ type DummyQueue struct {
|
|||||||
}
|
}
|
||||||
|
|
||||||
// Run does nothing
|
// Run does nothing
|
||||||
func (*DummyQueue) Run(_, _ func(context.Context, func())) {}
|
func (*DummyQueue) Run(_, _ func(func())) {}
|
||||||
|
|
||||||
// Push fakes a push of data to the queue
|
// Push fakes a push of data to the queue
|
||||||
func (*DummyQueue) Push(Data) error {
|
func (*DummyQueue) Push(Data) error {
|
||||||
@ -122,7 +122,7 @@ type Immediate struct {
|
|||||||
}
|
}
|
||||||
|
|
||||||
// Run does nothing
|
// Run does nothing
|
||||||
func (*Immediate) Run(_, _ func(context.Context, func())) {}
|
func (*Immediate) Run(_, _ func(func())) {}
|
||||||
|
|
||||||
// Push fakes a push of data to the queue
|
// Push fakes a push of data to the queue
|
||||||
func (q *Immediate) Push(data Data) error {
|
func (q *Immediate) Push(data Data) error {
|
||||||
|
@ -17,8 +17,9 @@ import (
|
|||||||
// ByteFIFOQueueConfiguration is the configuration for a ByteFIFOQueue
|
// ByteFIFOQueueConfiguration is the configuration for a ByteFIFOQueue
|
||||||
type ByteFIFOQueueConfiguration struct {
|
type ByteFIFOQueueConfiguration struct {
|
||||||
WorkerPoolConfiguration
|
WorkerPoolConfiguration
|
||||||
Workers int
|
Workers int
|
||||||
Name string
|
Name string
|
||||||
|
WaitOnEmpty bool
|
||||||
}
|
}
|
||||||
|
|
||||||
var _ Queue = &ByteFIFOQueue{}
|
var _ Queue = &ByteFIFOQueue{}
|
||||||
@ -26,14 +27,18 @@ var _ Queue = &ByteFIFOQueue{}
|
|||||||
// ByteFIFOQueue is a Queue formed from a ByteFIFO and WorkerPool
|
// ByteFIFOQueue is a Queue formed from a ByteFIFO and WorkerPool
|
||||||
type ByteFIFOQueue struct {
|
type ByteFIFOQueue struct {
|
||||||
*WorkerPool
|
*WorkerPool
|
||||||
byteFIFO ByteFIFO
|
byteFIFO ByteFIFO
|
||||||
typ Type
|
typ Type
|
||||||
closed chan struct{}
|
shutdownCtx context.Context
|
||||||
terminated chan struct{}
|
shutdownCtxCancel context.CancelFunc
|
||||||
exemplar interface{}
|
terminateCtx context.Context
|
||||||
workers int
|
terminateCtxCancel context.CancelFunc
|
||||||
name string
|
exemplar interface{}
|
||||||
lock sync.Mutex
|
workers int
|
||||||
|
name string
|
||||||
|
lock sync.Mutex
|
||||||
|
waitOnEmpty bool
|
||||||
|
pushed chan struct{}
|
||||||
}
|
}
|
||||||
|
|
||||||
// NewByteFIFOQueue creates a new ByteFIFOQueue
|
// NewByteFIFOQueue creates a new ByteFIFOQueue
|
||||||
@ -44,15 +49,22 @@ func NewByteFIFOQueue(typ Type, byteFIFO ByteFIFO, handle HandlerFunc, cfg, exem
|
|||||||
}
|
}
|
||||||
config := configInterface.(ByteFIFOQueueConfiguration)
|
config := configInterface.(ByteFIFOQueueConfiguration)
|
||||||
|
|
||||||
|
terminateCtx, terminateCtxCancel := context.WithCancel(context.Background())
|
||||||
|
shutdownCtx, shutdownCtxCancel := context.WithCancel(terminateCtx)
|
||||||
|
|
||||||
return &ByteFIFOQueue{
|
return &ByteFIFOQueue{
|
||||||
WorkerPool: NewWorkerPool(handle, config.WorkerPoolConfiguration),
|
WorkerPool: NewWorkerPool(handle, config.WorkerPoolConfiguration),
|
||||||
byteFIFO: byteFIFO,
|
byteFIFO: byteFIFO,
|
||||||
typ: typ,
|
typ: typ,
|
||||||
closed: make(chan struct{}),
|
shutdownCtx: shutdownCtx,
|
||||||
terminated: make(chan struct{}),
|
shutdownCtxCancel: shutdownCtxCancel,
|
||||||
exemplar: exemplar,
|
terminateCtx: terminateCtx,
|
||||||
workers: config.Workers,
|
terminateCtxCancel: terminateCtxCancel,
|
||||||
name: config.Name,
|
exemplar: exemplar,
|
||||||
|
workers: config.Workers,
|
||||||
|
name: config.Name,
|
||||||
|
waitOnEmpty: config.WaitOnEmpty,
|
||||||
|
pushed: make(chan struct{}, 1),
|
||||||
}, nil
|
}, nil
|
||||||
}
|
}
|
||||||
|
|
||||||
@ -76,7 +88,15 @@ func (q *ByteFIFOQueue) PushFunc(data Data, fn func() error) error {
|
|||||||
if err != nil {
|
if err != nil {
|
||||||
return err
|
return err
|
||||||
}
|
}
|
||||||
return q.byteFIFO.PushFunc(bs, fn)
|
if q.waitOnEmpty {
|
||||||
|
defer func() {
|
||||||
|
select {
|
||||||
|
case q.pushed <- struct{}{}:
|
||||||
|
default:
|
||||||
|
}
|
||||||
|
}()
|
||||||
|
}
|
||||||
|
return q.byteFIFO.PushFunc(q.terminateCtx, bs, fn)
|
||||||
}
|
}
|
||||||
|
|
||||||
// IsEmpty checks if the queue is empty
|
// IsEmpty checks if the queue is empty
|
||||||
@ -86,135 +106,160 @@ func (q *ByteFIFOQueue) IsEmpty() bool {
|
|||||||
if !q.WorkerPool.IsEmpty() {
|
if !q.WorkerPool.IsEmpty() {
|
||||||
return false
|
return false
|
||||||
}
|
}
|
||||||
return q.byteFIFO.Len() == 0
|
return q.byteFIFO.Len(q.terminateCtx) == 0
|
||||||
}
|
}
|
||||||
|
|
||||||
// Run runs the bytefifo queue
|
// Run runs the bytefifo queue
|
||||||
func (q *ByteFIFOQueue) Run(atShutdown, atTerminate func(context.Context, func())) {
|
func (q *ByteFIFOQueue) Run(atShutdown, atTerminate func(func())) {
|
||||||
atShutdown(context.Background(), q.Shutdown)
|
atShutdown(q.Shutdown)
|
||||||
atTerminate(context.Background(), q.Terminate)
|
atTerminate(q.Terminate)
|
||||||
log.Debug("%s: %s Starting", q.typ, q.name)
|
log.Debug("%s: %s Starting", q.typ, q.name)
|
||||||
|
|
||||||
go func() {
|
_ = q.AddWorkers(q.workers, 0)
|
||||||
_ = q.AddWorkers(q.workers, 0)
|
|
||||||
}()
|
|
||||||
|
|
||||||
go q.readToChan()
|
log.Trace("%s: %s Now running", q.typ, q.name)
|
||||||
|
q.readToChan()
|
||||||
|
|
||||||
log.Trace("%s: %s Waiting til closed", q.typ, q.name)
|
<-q.shutdownCtx.Done()
|
||||||
<-q.closed
|
|
||||||
log.Trace("%s: %s Waiting til done", q.typ, q.name)
|
log.Trace("%s: %s Waiting til done", q.typ, q.name)
|
||||||
q.Wait()
|
q.Wait()
|
||||||
|
|
||||||
log.Trace("%s: %s Waiting til cleaned", q.typ, q.name)
|
log.Trace("%s: %s Waiting til cleaned", q.typ, q.name)
|
||||||
ctx, cancel := context.WithCancel(context.Background())
|
q.CleanUp(q.terminateCtx)
|
||||||
atTerminate(ctx, cancel)
|
q.terminateCtxCancel()
|
||||||
q.CleanUp(ctx)
|
|
||||||
cancel()
|
|
||||||
}
|
}
|
||||||
|
|
||||||
|
const maxBackOffTime = time.Second * 3
|
||||||
|
|
||||||
func (q *ByteFIFOQueue) readToChan() {
|
func (q *ByteFIFOQueue) readToChan() {
|
||||||
// handle quick cancels
|
// handle quick cancels
|
||||||
select {
|
select {
|
||||||
case <-q.closed:
|
case <-q.shutdownCtx.Done():
|
||||||
// tell the pool to shutdown.
|
// tell the pool to shutdown.
|
||||||
q.cancel()
|
q.baseCtxCancel()
|
||||||
return
|
return
|
||||||
default:
|
default:
|
||||||
}
|
}
|
||||||
|
|
||||||
|
// Default backoff values
|
||||||
backOffTime := time.Millisecond * 100
|
backOffTime := time.Millisecond * 100
|
||||||
maxBackOffTime := time.Second * 3
|
|
||||||
|
loop:
|
||||||
for {
|
for {
|
||||||
success, resetBackoff := q.doPop()
|
err := q.doPop()
|
||||||
if resetBackoff {
|
if err == errQueueEmpty {
|
||||||
|
log.Trace("%s: %s Waiting on Empty", q.typ, q.name)
|
||||||
|
select {
|
||||||
|
case <-q.pushed:
|
||||||
|
// reset backOffTime
|
||||||
|
backOffTime = 100 * time.Millisecond
|
||||||
|
continue loop
|
||||||
|
case <-q.shutdownCtx.Done():
|
||||||
|
// Oops we've been shutdown whilst waiting
|
||||||
|
// Make sure the worker pool is shutdown too
|
||||||
|
q.baseCtxCancel()
|
||||||
|
return
|
||||||
|
}
|
||||||
|
}
|
||||||
|
|
||||||
|
// Reset the backOffTime if there is no error or an unmarshalError
|
||||||
|
if err == nil || err == errUnmarshal {
|
||||||
backOffTime = 100 * time.Millisecond
|
backOffTime = 100 * time.Millisecond
|
||||||
}
|
}
|
||||||
|
|
||||||
if success {
|
if err != nil {
|
||||||
|
// Need to Backoff
|
||||||
select {
|
select {
|
||||||
case <-q.closed:
|
case <-q.shutdownCtx.Done():
|
||||||
// tell the pool to shutdown.
|
// Oops we've been shutdown whilst backing off
|
||||||
q.cancel()
|
// Make sure the worker pool is shutdown too
|
||||||
return
|
q.baseCtxCancel()
|
||||||
default:
|
|
||||||
}
|
|
||||||
} else {
|
|
||||||
select {
|
|
||||||
case <-q.closed:
|
|
||||||
// tell the pool to shutdown.
|
|
||||||
q.cancel()
|
|
||||||
return
|
return
|
||||||
case <-time.After(backOffTime):
|
case <-time.After(backOffTime):
|
||||||
|
// OK we've waited - so backoff a bit
|
||||||
|
backOffTime += backOffTime / 2
|
||||||
|
if backOffTime > maxBackOffTime {
|
||||||
|
backOffTime = maxBackOffTime
|
||||||
|
}
|
||||||
|
continue loop
|
||||||
}
|
}
|
||||||
backOffTime += backOffTime / 2
|
}
|
||||||
if backOffTime > maxBackOffTime {
|
select {
|
||||||
backOffTime = maxBackOffTime
|
case <-q.shutdownCtx.Done():
|
||||||
}
|
// Oops we've been shutdown
|
||||||
|
// Make sure the worker pool is shutdown too
|
||||||
|
q.baseCtxCancel()
|
||||||
|
return
|
||||||
|
default:
|
||||||
|
continue loop
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
func (q *ByteFIFOQueue) doPop() (success, resetBackoff bool) {
|
var errQueueEmpty = fmt.Errorf("empty queue")
|
||||||
|
var errEmptyBytes = fmt.Errorf("empty bytes")
|
||||||
|
var errUnmarshal = fmt.Errorf("failed to unmarshal")
|
||||||
|
|
||||||
|
func (q *ByteFIFOQueue) doPop() error {
|
||||||
q.lock.Lock()
|
q.lock.Lock()
|
||||||
defer q.lock.Unlock()
|
defer q.lock.Unlock()
|
||||||
bs, err := q.byteFIFO.Pop()
|
bs, err := q.byteFIFO.Pop(q.shutdownCtx)
|
||||||
if err != nil {
|
if err != nil {
|
||||||
|
if err == context.Canceled {
|
||||||
|
q.baseCtxCancel()
|
||||||
|
return err
|
||||||
|
}
|
||||||
log.Error("%s: %s Error on Pop: %v", q.typ, q.name, err)
|
log.Error("%s: %s Error on Pop: %v", q.typ, q.name, err)
|
||||||
return
|
return err
|
||||||
}
|
}
|
||||||
if len(bs) == 0 {
|
if len(bs) == 0 {
|
||||||
return
|
if q.waitOnEmpty && q.byteFIFO.Len(q.shutdownCtx) == 0 {
|
||||||
|
return errQueueEmpty
|
||||||
|
}
|
||||||
|
return errEmptyBytes
|
||||||
}
|
}
|
||||||
|
|
||||||
resetBackoff = true
|
|
||||||
|
|
||||||
data, err := unmarshalAs(bs, q.exemplar)
|
data, err := unmarshalAs(bs, q.exemplar)
|
||||||
if err != nil {
|
if err != nil {
|
||||||
log.Error("%s: %s Failed to unmarshal with error: %v", q.typ, q.name, err)
|
log.Error("%s: %s Failed to unmarshal with error: %v", q.typ, q.name, err)
|
||||||
return
|
return errUnmarshal
|
||||||
}
|
}
|
||||||
|
|
||||||
log.Trace("%s %s: Task found: %#v", q.typ, q.name, data)
|
log.Trace("%s %s: Task found: %#v", q.typ, q.name, data)
|
||||||
q.WorkerPool.Push(data)
|
q.WorkerPool.Push(data)
|
||||||
success = true
|
return nil
|
||||||
return
|
|
||||||
}
|
}
|
||||||
|
|
||||||
// Shutdown processing from this queue
|
// Shutdown processing from this queue
|
||||||
func (q *ByteFIFOQueue) Shutdown() {
|
func (q *ByteFIFOQueue) Shutdown() {
|
||||||
log.Trace("%s: %s Shutting down", q.typ, q.name)
|
log.Trace("%s: %s Shutting down", q.typ, q.name)
|
||||||
q.lock.Lock()
|
|
||||||
select {
|
select {
|
||||||
case <-q.closed:
|
case <-q.shutdownCtx.Done():
|
||||||
|
return
|
||||||
default:
|
default:
|
||||||
close(q.closed)
|
|
||||||
}
|
}
|
||||||
q.lock.Unlock()
|
q.shutdownCtxCancel()
|
||||||
log.Debug("%s: %s Shutdown", q.typ, q.name)
|
log.Debug("%s: %s Shutdown", q.typ, q.name)
|
||||||
}
|
}
|
||||||
|
|
||||||
// IsShutdown returns a channel which is closed when this Queue is shutdown
|
// IsShutdown returns a channel which is closed when this Queue is shutdown
|
||||||
func (q *ByteFIFOQueue) IsShutdown() <-chan struct{} {
|
func (q *ByteFIFOQueue) IsShutdown() <-chan struct{} {
|
||||||
return q.closed
|
return q.shutdownCtx.Done()
|
||||||
}
|
}
|
||||||
|
|
||||||
// Terminate this queue and close the queue
|
// Terminate this queue and close the queue
|
||||||
func (q *ByteFIFOQueue) Terminate() {
|
func (q *ByteFIFOQueue) Terminate() {
|
||||||
log.Trace("%s: %s Terminating", q.typ, q.name)
|
log.Trace("%s: %s Terminating", q.typ, q.name)
|
||||||
q.Shutdown()
|
q.Shutdown()
|
||||||
q.lock.Lock()
|
|
||||||
select {
|
select {
|
||||||
case <-q.terminated:
|
case <-q.terminateCtx.Done():
|
||||||
q.lock.Unlock()
|
|
||||||
return
|
return
|
||||||
default:
|
default:
|
||||||
}
|
}
|
||||||
close(q.terminated)
|
|
||||||
q.lock.Unlock()
|
|
||||||
if log.IsDebug() {
|
if log.IsDebug() {
|
||||||
log.Debug("%s: %s Closing with %d tasks left in queue", q.typ, q.name, q.byteFIFO.Len())
|
log.Debug("%s: %s Closing with %d tasks left in queue", q.typ, q.name, q.byteFIFO.Len(q.terminateCtx))
|
||||||
}
|
}
|
||||||
|
q.terminateCtxCancel()
|
||||||
if err := q.byteFIFO.Close(); err != nil {
|
if err := q.byteFIFO.Close(); err != nil {
|
||||||
log.Error("Error whilst closing internal byte fifo in %s: %s: %v", q.typ, q.name, err)
|
log.Error("Error whilst closing internal byte fifo in %s: %s: %v", q.typ, q.name, err)
|
||||||
}
|
}
|
||||||
@ -223,7 +268,7 @@ func (q *ByteFIFOQueue) Terminate() {
|
|||||||
|
|
||||||
// IsTerminated returns a channel which is closed when this Queue is terminated
|
// IsTerminated returns a channel which is closed when this Queue is terminated
|
||||||
func (q *ByteFIFOQueue) IsTerminated() <-chan struct{} {
|
func (q *ByteFIFOQueue) IsTerminated() <-chan struct{} {
|
||||||
return q.terminated
|
return q.terminateCtx.Done()
|
||||||
}
|
}
|
||||||
|
|
||||||
var _ UniqueQueue = &ByteFIFOUniqueQueue{}
|
var _ UniqueQueue = &ByteFIFOUniqueQueue{}
|
||||||
@ -240,17 +285,21 @@ func NewByteFIFOUniqueQueue(typ Type, byteFIFO UniqueByteFIFO, handle HandlerFun
|
|||||||
return nil, err
|
return nil, err
|
||||||
}
|
}
|
||||||
config := configInterface.(ByteFIFOQueueConfiguration)
|
config := configInterface.(ByteFIFOQueueConfiguration)
|
||||||
|
terminateCtx, terminateCtxCancel := context.WithCancel(context.Background())
|
||||||
|
shutdownCtx, shutdownCtxCancel := context.WithCancel(terminateCtx)
|
||||||
|
|
||||||
return &ByteFIFOUniqueQueue{
|
return &ByteFIFOUniqueQueue{
|
||||||
ByteFIFOQueue: ByteFIFOQueue{
|
ByteFIFOQueue: ByteFIFOQueue{
|
||||||
WorkerPool: NewWorkerPool(handle, config.WorkerPoolConfiguration),
|
WorkerPool: NewWorkerPool(handle, config.WorkerPoolConfiguration),
|
||||||
byteFIFO: byteFIFO,
|
byteFIFO: byteFIFO,
|
||||||
typ: typ,
|
typ: typ,
|
||||||
closed: make(chan struct{}),
|
shutdownCtx: shutdownCtx,
|
||||||
terminated: make(chan struct{}),
|
shutdownCtxCancel: shutdownCtxCancel,
|
||||||
exemplar: exemplar,
|
terminateCtx: terminateCtx,
|
||||||
workers: config.Workers,
|
terminateCtxCancel: terminateCtxCancel,
|
||||||
name: config.Name,
|
exemplar: exemplar,
|
||||||
|
workers: config.Workers,
|
||||||
|
name: config.Name,
|
||||||
},
|
},
|
||||||
}, nil
|
}, nil
|
||||||
}
|
}
|
||||||
@ -265,5 +314,5 @@ func (q *ByteFIFOUniqueQueue) Has(data Data) (bool, error) {
|
|||||||
if err != nil {
|
if err != nil {
|
||||||
return false, err
|
return false, err
|
||||||
}
|
}
|
||||||
return q.byteFIFO.(UniqueByteFIFO).Has(bs)
|
return q.byteFIFO.(UniqueByteFIFO).Has(q.terminateCtx, bs)
|
||||||
}
|
}
|
||||||
|
@ -27,9 +27,13 @@ type ChannelQueueConfiguration struct {
|
|||||||
// It is basically a very thin wrapper around a WorkerPool
|
// It is basically a very thin wrapper around a WorkerPool
|
||||||
type ChannelQueue struct {
|
type ChannelQueue struct {
|
||||||
*WorkerPool
|
*WorkerPool
|
||||||
exemplar interface{}
|
shutdownCtx context.Context
|
||||||
workers int
|
shutdownCtxCancel context.CancelFunc
|
||||||
name string
|
terminateCtx context.Context
|
||||||
|
terminateCtxCancel context.CancelFunc
|
||||||
|
exemplar interface{}
|
||||||
|
workers int
|
||||||
|
name string
|
||||||
}
|
}
|
||||||
|
|
||||||
// NewChannelQueue creates a memory channel queue
|
// NewChannelQueue creates a memory channel queue
|
||||||
@ -42,28 +46,30 @@ func NewChannelQueue(handle HandlerFunc, cfg, exemplar interface{}) (Queue, erro
|
|||||||
if config.BatchLength == 0 {
|
if config.BatchLength == 0 {
|
||||||
config.BatchLength = 1
|
config.BatchLength = 1
|
||||||
}
|
}
|
||||||
|
|
||||||
|
terminateCtx, terminateCtxCancel := context.WithCancel(context.Background())
|
||||||
|
shutdownCtx, shutdownCtxCancel := context.WithCancel(terminateCtx)
|
||||||
|
|
||||||
queue := &ChannelQueue{
|
queue := &ChannelQueue{
|
||||||
WorkerPool: NewWorkerPool(handle, config.WorkerPoolConfiguration),
|
WorkerPool: NewWorkerPool(handle, config.WorkerPoolConfiguration),
|
||||||
exemplar: exemplar,
|
shutdownCtx: shutdownCtx,
|
||||||
workers: config.Workers,
|
shutdownCtxCancel: shutdownCtxCancel,
|
||||||
name: config.Name,
|
terminateCtx: terminateCtx,
|
||||||
|
terminateCtxCancel: terminateCtxCancel,
|
||||||
|
exemplar: exemplar,
|
||||||
|
workers: config.Workers,
|
||||||
|
name: config.Name,
|
||||||
}
|
}
|
||||||
queue.qid = GetManager().Add(queue, ChannelQueueType, config, exemplar)
|
queue.qid = GetManager().Add(queue, ChannelQueueType, config, exemplar)
|
||||||
return queue, nil
|
return queue, nil
|
||||||
}
|
}
|
||||||
|
|
||||||
// Run starts to run the queue
|
// Run starts to run the queue
|
||||||
func (q *ChannelQueue) Run(atShutdown, atTerminate func(context.Context, func())) {
|
func (q *ChannelQueue) Run(atShutdown, atTerminate func(func())) {
|
||||||
atShutdown(context.Background(), func() {
|
atShutdown(q.Shutdown)
|
||||||
log.Warn("ChannelQueue: %s is not shutdownable!", q.name)
|
atTerminate(q.Terminate)
|
||||||
})
|
|
||||||
atTerminate(context.Background(), func() {
|
|
||||||
log.Warn("ChannelQueue: %s is not terminatable!", q.name)
|
|
||||||
})
|
|
||||||
log.Debug("ChannelQueue: %s Starting", q.name)
|
log.Debug("ChannelQueue: %s Starting", q.name)
|
||||||
go func() {
|
_ = q.AddWorkers(q.workers, 0)
|
||||||
_ = q.AddWorkers(q.workers, 0)
|
|
||||||
}()
|
|
||||||
}
|
}
|
||||||
|
|
||||||
// Push will push data into the queue
|
// Push will push data into the queue
|
||||||
@ -75,6 +81,42 @@ func (q *ChannelQueue) Push(data Data) error {
|
|||||||
return nil
|
return nil
|
||||||
}
|
}
|
||||||
|
|
||||||
|
// Shutdown processing from this queue
|
||||||
|
func (q *ChannelQueue) Shutdown() {
|
||||||
|
q.lock.Lock()
|
||||||
|
defer q.lock.Unlock()
|
||||||
|
select {
|
||||||
|
case <-q.shutdownCtx.Done():
|
||||||
|
log.Trace("ChannelQueue: %s Already Shutting down", q.name)
|
||||||
|
return
|
||||||
|
default:
|
||||||
|
}
|
||||||
|
log.Trace("ChannelQueue: %s Shutting down", q.name)
|
||||||
|
go func() {
|
||||||
|
log.Trace("ChannelQueue: %s Flushing", q.name)
|
||||||
|
if err := q.FlushWithContext(q.terminateCtx); err != nil {
|
||||||
|
log.Warn("ChannelQueue: %s Terminated before completed flushing", q.name)
|
||||||
|
return
|
||||||
|
}
|
||||||
|
log.Debug("ChannelQueue: %s Flushed", q.name)
|
||||||
|
}()
|
||||||
|
q.shutdownCtxCancel()
|
||||||
|
log.Debug("ChannelQueue: %s Shutdown", q.name)
|
||||||
|
}
|
||||||
|
|
||||||
|
// Terminate this queue and close the queue
|
||||||
|
func (q *ChannelQueue) Terminate() {
|
||||||
|
log.Trace("ChannelQueue: %s Terminating", q.name)
|
||||||
|
q.Shutdown()
|
||||||
|
select {
|
||||||
|
case <-q.terminateCtx.Done():
|
||||||
|
return
|
||||||
|
default:
|
||||||
|
}
|
||||||
|
q.terminateCtxCancel()
|
||||||
|
log.Debug("ChannelQueue: %s Terminated", q.name)
|
||||||
|
}
|
||||||
|
|
||||||
// Name returns the name of this queue
|
// Name returns the name of this queue
|
||||||
func (q *ChannelQueue) Name() string {
|
func (q *ChannelQueue) Name() string {
|
||||||
return q.name
|
return q.name
|
||||||
|
@ -5,7 +5,6 @@
|
|||||||
package queue
|
package queue
|
||||||
|
|
||||||
import (
|
import (
|
||||||
"context"
|
|
||||||
"testing"
|
"testing"
|
||||||
"time"
|
"time"
|
||||||
|
|
||||||
@ -21,7 +20,7 @@ func TestChannelQueue(t *testing.T) {
|
|||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
nilFn := func(_ context.Context, _ func()) {}
|
nilFn := func(_ func()) {}
|
||||||
|
|
||||||
queue, err := NewChannelQueue(handle,
|
queue, err := NewChannelQueue(handle,
|
||||||
ChannelQueueConfiguration{
|
ChannelQueueConfiguration{
|
||||||
@ -61,7 +60,7 @@ func TestChannelQueue_Batch(t *testing.T) {
|
|||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
nilFn := func(_ context.Context, _ func()) {}
|
nilFn := func(_ func()) {}
|
||||||
|
|
||||||
queue, err := NewChannelQueue(handle,
|
queue, err := NewChannelQueue(handle,
|
||||||
ChannelQueueConfiguration{
|
ChannelQueueConfiguration{
|
||||||
|
@ -5,6 +5,8 @@
|
|||||||
package queue
|
package queue
|
||||||
|
|
||||||
import (
|
import (
|
||||||
|
"context"
|
||||||
|
|
||||||
"code.gitea.io/gitea/modules/nosql"
|
"code.gitea.io/gitea/modules/nosql"
|
||||||
|
|
||||||
"gitea.com/lunny/levelqueue"
|
"gitea.com/lunny/levelqueue"
|
||||||
@ -37,6 +39,7 @@ func NewLevelQueue(handle HandlerFunc, cfg, exemplar interface{}) (Queue, error)
|
|||||||
if len(config.ConnectionString) == 0 {
|
if len(config.ConnectionString) == 0 {
|
||||||
config.ConnectionString = config.DataDir
|
config.ConnectionString = config.DataDir
|
||||||
}
|
}
|
||||||
|
config.WaitOnEmpty = true
|
||||||
|
|
||||||
byteFIFO, err := NewLevelQueueByteFIFO(config.ConnectionString, config.QueueName)
|
byteFIFO, err := NewLevelQueueByteFIFO(config.ConnectionString, config.QueueName)
|
||||||
if err != nil {
|
if err != nil {
|
||||||
@ -82,7 +85,7 @@ func NewLevelQueueByteFIFO(connection, prefix string) (*LevelQueueByteFIFO, erro
|
|||||||
}
|
}
|
||||||
|
|
||||||
// PushFunc will push data into the fifo
|
// PushFunc will push data into the fifo
|
||||||
func (fifo *LevelQueueByteFIFO) PushFunc(data []byte, fn func() error) error {
|
func (fifo *LevelQueueByteFIFO) PushFunc(ctx context.Context, data []byte, fn func() error) error {
|
||||||
if fn != nil {
|
if fn != nil {
|
||||||
if err := fn(); err != nil {
|
if err := fn(); err != nil {
|
||||||
return err
|
return err
|
||||||
@ -92,7 +95,7 @@ func (fifo *LevelQueueByteFIFO) PushFunc(data []byte, fn func() error) error {
|
|||||||
}
|
}
|
||||||
|
|
||||||
// Pop pops data from the start of the fifo
|
// Pop pops data from the start of the fifo
|
||||||
func (fifo *LevelQueueByteFIFO) Pop() ([]byte, error) {
|
func (fifo *LevelQueueByteFIFO) Pop(ctx context.Context) ([]byte, error) {
|
||||||
data, err := fifo.internal.RPop()
|
data, err := fifo.internal.RPop()
|
||||||
if err != nil && err != levelqueue.ErrNotFound {
|
if err != nil && err != levelqueue.ErrNotFound {
|
||||||
return nil, err
|
return nil, err
|
||||||
@ -108,7 +111,7 @@ func (fifo *LevelQueueByteFIFO) Close() error {
|
|||||||
}
|
}
|
||||||
|
|
||||||
// Len returns the length of the fifo
|
// Len returns the length of the fifo
|
||||||
func (fifo *LevelQueueByteFIFO) Len() int64 {
|
func (fifo *LevelQueueByteFIFO) Len(ctx context.Context) int64 {
|
||||||
return fifo.internal.Len()
|
return fifo.internal.Len()
|
||||||
}
|
}
|
||||||
|
|
||||||
|
@ -133,8 +133,9 @@ func (q *PersistableChannelQueue) Push(data Data) error {
|
|||||||
}
|
}
|
||||||
|
|
||||||
// Run starts to run the queue
|
// Run starts to run the queue
|
||||||
func (q *PersistableChannelQueue) Run(atShutdown, atTerminate func(context.Context, func())) {
|
func (q *PersistableChannelQueue) Run(atShutdown, atTerminate func(func())) {
|
||||||
log.Debug("PersistableChannelQueue: %s Starting", q.delayedStarter.name)
|
log.Debug("PersistableChannelQueue: %s Starting", q.delayedStarter.name)
|
||||||
|
_ = q.channelQueue.AddWorkers(q.channelQueue.workers, 0)
|
||||||
|
|
||||||
q.lock.Lock()
|
q.lock.Lock()
|
||||||
if q.internal == nil {
|
if q.internal == nil {
|
||||||
@ -147,34 +148,32 @@ func (q *PersistableChannelQueue) Run(atShutdown, atTerminate func(context.Conte
|
|||||||
} else {
|
} else {
|
||||||
q.lock.Unlock()
|
q.lock.Unlock()
|
||||||
}
|
}
|
||||||
atShutdown(context.Background(), q.Shutdown)
|
atShutdown(q.Shutdown)
|
||||||
atTerminate(context.Background(), q.Terminate)
|
atTerminate(q.Terminate)
|
||||||
|
|
||||||
// Just run the level queue - we shut it down later
|
if lq, ok := q.internal.(*LevelQueue); ok && lq.byteFIFO.Len(lq.shutdownCtx) != 0 {
|
||||||
go q.internal.Run(func(_ context.Context, _ func()) {}, func(_ context.Context, _ func()) {})
|
// Just run the level queue - we shut it down once it's flushed
|
||||||
|
go q.internal.Run(func(_ func()) {}, func(_ func()) {})
|
||||||
|
go func() {
|
||||||
|
for !q.IsEmpty() {
|
||||||
|
_ = q.internal.Flush(0)
|
||||||
|
select {
|
||||||
|
case <-time.After(100 * time.Millisecond):
|
||||||
|
case <-q.internal.(*LevelQueue).shutdownCtx.Done():
|
||||||
|
log.Warn("LevelQueue: %s shut down before completely flushed", q.internal.(*LevelQueue).Name())
|
||||||
|
return
|
||||||
|
}
|
||||||
|
}
|
||||||
|
log.Debug("LevelQueue: %s flushed so shutting down", q.internal.(*LevelQueue).Name())
|
||||||
|
q.internal.(*LevelQueue).Shutdown()
|
||||||
|
GetManager().Remove(q.internal.(*LevelQueue).qid)
|
||||||
|
}()
|
||||||
|
} else {
|
||||||
|
log.Debug("PersistableChannelQueue: %s Skipping running the empty level queue", q.delayedStarter.name)
|
||||||
|
q.internal.(*LevelQueue).Shutdown()
|
||||||
|
GetManager().Remove(q.internal.(*LevelQueue).qid)
|
||||||
|
}
|
||||||
|
|
||||||
go func() {
|
|
||||||
_ = q.channelQueue.AddWorkers(q.channelQueue.workers, 0)
|
|
||||||
}()
|
|
||||||
|
|
||||||
log.Trace("PersistableChannelQueue: %s Waiting til closed", q.delayedStarter.name)
|
|
||||||
<-q.closed
|
|
||||||
log.Trace("PersistableChannelQueue: %s Cancelling pools", q.delayedStarter.name)
|
|
||||||
q.channelQueue.cancel()
|
|
||||||
q.internal.(*LevelQueue).cancel()
|
|
||||||
log.Trace("PersistableChannelQueue: %s Waiting til done", q.delayedStarter.name)
|
|
||||||
q.channelQueue.Wait()
|
|
||||||
q.internal.(*LevelQueue).Wait()
|
|
||||||
// Redirect all remaining data in the chan to the internal channel
|
|
||||||
go func() {
|
|
||||||
log.Trace("PersistableChannelQueue: %s Redirecting remaining data", q.delayedStarter.name)
|
|
||||||
for data := range q.channelQueue.dataChan {
|
|
||||||
_ = q.internal.Push(data)
|
|
||||||
atomic.AddInt64(&q.channelQueue.numInQueue, -1)
|
|
||||||
}
|
|
||||||
log.Trace("PersistableChannelQueue: %s Done Redirecting remaining data", q.delayedStarter.name)
|
|
||||||
}()
|
|
||||||
log.Trace("PersistableChannelQueue: %s Done main loop", q.delayedStarter.name)
|
|
||||||
}
|
}
|
||||||
|
|
||||||
// Flush flushes the queue and blocks till the queue is empty
|
// Flush flushes the queue and blocks till the queue is empty
|
||||||
@ -232,16 +231,37 @@ func (q *PersistableChannelQueue) IsEmpty() bool {
|
|||||||
func (q *PersistableChannelQueue) Shutdown() {
|
func (q *PersistableChannelQueue) Shutdown() {
|
||||||
log.Trace("PersistableChannelQueue: %s Shutting down", q.delayedStarter.name)
|
log.Trace("PersistableChannelQueue: %s Shutting down", q.delayedStarter.name)
|
||||||
q.lock.Lock()
|
q.lock.Lock()
|
||||||
defer q.lock.Unlock()
|
|
||||||
select {
|
select {
|
||||||
case <-q.closed:
|
case <-q.closed:
|
||||||
|
q.lock.Unlock()
|
||||||
|
return
|
||||||
default:
|
default:
|
||||||
if q.internal != nil {
|
|
||||||
q.internal.(*LevelQueue).Shutdown()
|
|
||||||
}
|
|
||||||
close(q.closed)
|
|
||||||
log.Debug("PersistableChannelQueue: %s Shutdown", q.delayedStarter.name)
|
|
||||||
}
|
}
|
||||||
|
q.channelQueue.Shutdown()
|
||||||
|
if q.internal != nil {
|
||||||
|
q.internal.(*LevelQueue).Shutdown()
|
||||||
|
}
|
||||||
|
close(q.closed)
|
||||||
|
q.lock.Unlock()
|
||||||
|
|
||||||
|
log.Trace("PersistableChannelQueue: %s Cancelling pools", q.delayedStarter.name)
|
||||||
|
q.channelQueue.baseCtxCancel()
|
||||||
|
q.internal.(*LevelQueue).baseCtxCancel()
|
||||||
|
log.Trace("PersistableChannelQueue: %s Waiting til done", q.delayedStarter.name)
|
||||||
|
q.channelQueue.Wait()
|
||||||
|
q.internal.(*LevelQueue).Wait()
|
||||||
|
// Redirect all remaining data in the chan to the internal channel
|
||||||
|
go func() {
|
||||||
|
log.Trace("PersistableChannelQueue: %s Redirecting remaining data", q.delayedStarter.name)
|
||||||
|
for data := range q.channelQueue.dataChan {
|
||||||
|
_ = q.internal.Push(data)
|
||||||
|
atomic.AddInt64(&q.channelQueue.numInQueue, -1)
|
||||||
|
}
|
||||||
|
log.Trace("PersistableChannelQueue: %s Done Redirecting remaining data", q.delayedStarter.name)
|
||||||
|
}()
|
||||||
|
|
||||||
|
log.Debug("PersistableChannelQueue: %s Shutdown", q.delayedStarter.name)
|
||||||
}
|
}
|
||||||
|
|
||||||
// Terminate this queue and close the queue
|
// Terminate this queue and close the queue
|
||||||
@ -250,6 +270,7 @@ func (q *PersistableChannelQueue) Terminate() {
|
|||||||
q.Shutdown()
|
q.Shutdown()
|
||||||
q.lock.Lock()
|
q.lock.Lock()
|
||||||
defer q.lock.Unlock()
|
defer q.lock.Unlock()
|
||||||
|
q.channelQueue.Terminate()
|
||||||
if q.internal != nil {
|
if q.internal != nil {
|
||||||
q.internal.(*LevelQueue).Terminate()
|
q.internal.(*LevelQueue).Terminate()
|
||||||
}
|
}
|
||||||
|
@ -5,10 +5,8 @@
|
|||||||
package queue
|
package queue
|
||||||
|
|
||||||
import (
|
import (
|
||||||
"context"
|
|
||||||
"io/ioutil"
|
"io/ioutil"
|
||||||
"testing"
|
"testing"
|
||||||
"time"
|
|
||||||
|
|
||||||
"code.gitea.io/gitea/modules/util"
|
"code.gitea.io/gitea/modules/util"
|
||||||
"github.com/stretchr/testify/assert"
|
"github.com/stretchr/testify/assert"
|
||||||
@ -32,17 +30,19 @@ func TestPersistableChannelQueue(t *testing.T) {
|
|||||||
defer util.RemoveAll(tmpDir)
|
defer util.RemoveAll(tmpDir)
|
||||||
|
|
||||||
queue, err := NewPersistableChannelQueue(handle, PersistableChannelQueueConfiguration{
|
queue, err := NewPersistableChannelQueue(handle, PersistableChannelQueueConfiguration{
|
||||||
DataDir: tmpDir,
|
DataDir: tmpDir,
|
||||||
BatchLength: 2,
|
BatchLength: 2,
|
||||||
QueueLength: 20,
|
QueueLength: 20,
|
||||||
Workers: 1,
|
Workers: 1,
|
||||||
MaxWorkers: 10,
|
BoostWorkers: 0,
|
||||||
|
MaxWorkers: 10,
|
||||||
|
Name: "first",
|
||||||
}, &testData{})
|
}, &testData{})
|
||||||
assert.NoError(t, err)
|
assert.NoError(t, err)
|
||||||
|
|
||||||
go queue.Run(func(_ context.Context, shutdown func()) {
|
go queue.Run(func(shutdown func()) {
|
||||||
queueShutdown = append(queueShutdown, shutdown)
|
queueShutdown = append(queueShutdown, shutdown)
|
||||||
}, func(_ context.Context, terminate func()) {
|
}, func(terminate func()) {
|
||||||
queueTerminate = append(queueTerminate, terminate)
|
queueTerminate = append(queueTerminate, terminate)
|
||||||
})
|
})
|
||||||
|
|
||||||
@ -64,13 +64,18 @@ func TestPersistableChannelQueue(t *testing.T) {
|
|||||||
assert.Equal(t, test2.TestString, result2.TestString)
|
assert.Equal(t, test2.TestString, result2.TestString)
|
||||||
assert.Equal(t, test2.TestInt, result2.TestInt)
|
assert.Equal(t, test2.TestInt, result2.TestInt)
|
||||||
|
|
||||||
|
// test1 is a testData not a *testData so will be rejected
|
||||||
err = queue.Push(test1)
|
err = queue.Push(test1)
|
||||||
assert.Error(t, err)
|
assert.Error(t, err)
|
||||||
|
|
||||||
|
// Now shutdown the queue
|
||||||
for _, callback := range queueShutdown {
|
for _, callback := range queueShutdown {
|
||||||
callback()
|
callback()
|
||||||
}
|
}
|
||||||
time.Sleep(200 * time.Millisecond)
|
|
||||||
|
// Wait til it is closed
|
||||||
|
<-queue.(*PersistableChannelQueue).closed
|
||||||
|
|
||||||
err = queue.Push(&test1)
|
err = queue.Push(&test1)
|
||||||
assert.NoError(t, err)
|
assert.NoError(t, err)
|
||||||
err = queue.Push(&test2)
|
err = queue.Push(&test2)
|
||||||
@ -80,23 +85,33 @@ func TestPersistableChannelQueue(t *testing.T) {
|
|||||||
assert.Fail(t, "Handler processing should have stopped")
|
assert.Fail(t, "Handler processing should have stopped")
|
||||||
default:
|
default:
|
||||||
}
|
}
|
||||||
|
|
||||||
|
// terminate the queue
|
||||||
for _, callback := range queueTerminate {
|
for _, callback := range queueTerminate {
|
||||||
callback()
|
callback()
|
||||||
}
|
}
|
||||||
|
|
||||||
|
select {
|
||||||
|
case <-handleChan:
|
||||||
|
assert.Fail(t, "Handler processing should have stopped")
|
||||||
|
default:
|
||||||
|
}
|
||||||
|
|
||||||
// Reopen queue
|
// Reopen queue
|
||||||
queue, err = NewPersistableChannelQueue(handle, PersistableChannelQueueConfiguration{
|
queue, err = NewPersistableChannelQueue(handle, PersistableChannelQueueConfiguration{
|
||||||
DataDir: tmpDir,
|
DataDir: tmpDir,
|
||||||
BatchLength: 2,
|
BatchLength: 2,
|
||||||
QueueLength: 20,
|
QueueLength: 20,
|
||||||
Workers: 1,
|
Workers: 1,
|
||||||
MaxWorkers: 10,
|
BoostWorkers: 0,
|
||||||
|
MaxWorkers: 10,
|
||||||
|
Name: "second",
|
||||||
}, &testData{})
|
}, &testData{})
|
||||||
assert.NoError(t, err)
|
assert.NoError(t, err)
|
||||||
|
|
||||||
go queue.Run(func(_ context.Context, shutdown func()) {
|
go queue.Run(func(shutdown func()) {
|
||||||
queueShutdown = append(queueShutdown, shutdown)
|
queueShutdown = append(queueShutdown, shutdown)
|
||||||
}, func(_ context.Context, terminate func()) {
|
}, func(terminate func()) {
|
||||||
queueTerminate = append(queueTerminate, terminate)
|
queueTerminate = append(queueTerminate, terminate)
|
||||||
})
|
})
|
||||||
|
|
||||||
|
@ -5,7 +5,6 @@
|
|||||||
package queue
|
package queue
|
||||||
|
|
||||||
import (
|
import (
|
||||||
"context"
|
|
||||||
"io/ioutil"
|
"io/ioutil"
|
||||||
"sync"
|
"sync"
|
||||||
"testing"
|
"testing"
|
||||||
@ -49,11 +48,11 @@ func TestLevelQueue(t *testing.T) {
|
|||||||
}, &testData{})
|
}, &testData{})
|
||||||
assert.NoError(t, err)
|
assert.NoError(t, err)
|
||||||
|
|
||||||
go queue.Run(func(_ context.Context, shutdown func()) {
|
go queue.Run(func(shutdown func()) {
|
||||||
lock.Lock()
|
lock.Lock()
|
||||||
queueShutdown = append(queueShutdown, shutdown)
|
queueShutdown = append(queueShutdown, shutdown)
|
||||||
lock.Unlock()
|
lock.Unlock()
|
||||||
}, func(_ context.Context, terminate func()) {
|
}, func(terminate func()) {
|
||||||
lock.Lock()
|
lock.Lock()
|
||||||
queueTerminate = append(queueTerminate, terminate)
|
queueTerminate = append(queueTerminate, terminate)
|
||||||
lock.Unlock()
|
lock.Unlock()
|
||||||
@ -123,11 +122,11 @@ func TestLevelQueue(t *testing.T) {
|
|||||||
}, &testData{})
|
}, &testData{})
|
||||||
assert.NoError(t, err)
|
assert.NoError(t, err)
|
||||||
|
|
||||||
go queue.Run(func(_ context.Context, shutdown func()) {
|
go queue.Run(func(shutdown func()) {
|
||||||
lock.Lock()
|
lock.Lock()
|
||||||
queueShutdown = append(queueShutdown, shutdown)
|
queueShutdown = append(queueShutdown, shutdown)
|
||||||
lock.Unlock()
|
lock.Unlock()
|
||||||
}, func(_ context.Context, terminate func()) {
|
}, func(terminate func()) {
|
||||||
lock.Lock()
|
lock.Lock()
|
||||||
queueTerminate = append(queueTerminate, terminate)
|
queueTerminate = append(queueTerminate, terminate)
|
||||||
lock.Unlock()
|
lock.Unlock()
|
||||||
|
@ -6,7 +6,6 @@ package queue
|
|||||||
|
|
||||||
import (
|
import (
|
||||||
"context"
|
"context"
|
||||||
"fmt"
|
|
||||||
|
|
||||||
"code.gitea.io/gitea/modules/graceful"
|
"code.gitea.io/gitea/modules/graceful"
|
||||||
"code.gitea.io/gitea/modules/log"
|
"code.gitea.io/gitea/modules/log"
|
||||||
@ -47,8 +46,6 @@ func NewRedisQueue(handle HandlerFunc, cfg, exemplar interface{}) (Queue, error)
|
|||||||
return nil, err
|
return nil, err
|
||||||
}
|
}
|
||||||
|
|
||||||
byteFIFO.ctx = graceful.NewChannelContext(byteFIFOQueue.IsTerminated(), fmt.Errorf("queue has been terminated"))
|
|
||||||
|
|
||||||
queue := &RedisQueue{
|
queue := &RedisQueue{
|
||||||
ByteFIFOQueue: byteFIFOQueue,
|
ByteFIFOQueue: byteFIFOQueue,
|
||||||
}
|
}
|
||||||
@ -73,8 +70,8 @@ var _ ByteFIFO = &RedisByteFIFO{}
|
|||||||
|
|
||||||
// RedisByteFIFO represents a ByteFIFO formed from a redisClient
|
// RedisByteFIFO represents a ByteFIFO formed from a redisClient
|
||||||
type RedisByteFIFO struct {
|
type RedisByteFIFO struct {
|
||||||
ctx context.Context
|
client redisClient
|
||||||
client redisClient
|
|
||||||
queueName string
|
queueName string
|
||||||
}
|
}
|
||||||
|
|
||||||
@ -89,7 +86,6 @@ func NewRedisByteFIFO(config RedisByteFIFOConfiguration) (*RedisByteFIFO, error)
|
|||||||
fifo := &RedisByteFIFO{
|
fifo := &RedisByteFIFO{
|
||||||
queueName: config.QueueName,
|
queueName: config.QueueName,
|
||||||
}
|
}
|
||||||
fifo.ctx = graceful.GetManager().TerminateContext()
|
|
||||||
fifo.client = nosql.GetManager().GetRedisClient(config.ConnectionString)
|
fifo.client = nosql.GetManager().GetRedisClient(config.ConnectionString)
|
||||||
if err := fifo.client.Ping(graceful.GetManager().ShutdownContext()).Err(); err != nil {
|
if err := fifo.client.Ping(graceful.GetManager().ShutdownContext()).Err(); err != nil {
|
||||||
return nil, err
|
return nil, err
|
||||||
@ -98,18 +94,18 @@ func NewRedisByteFIFO(config RedisByteFIFOConfiguration) (*RedisByteFIFO, error)
|
|||||||
}
|
}
|
||||||
|
|
||||||
// PushFunc pushes data to the end of the fifo and calls the callback if it is added
|
// PushFunc pushes data to the end of the fifo and calls the callback if it is added
|
||||||
func (fifo *RedisByteFIFO) PushFunc(data []byte, fn func() error) error {
|
func (fifo *RedisByteFIFO) PushFunc(ctx context.Context, data []byte, fn func() error) error {
|
||||||
if fn != nil {
|
if fn != nil {
|
||||||
if err := fn(); err != nil {
|
if err := fn(); err != nil {
|
||||||
return err
|
return err
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
return fifo.client.RPush(fifo.ctx, fifo.queueName, data).Err()
|
return fifo.client.RPush(ctx, fifo.queueName, data).Err()
|
||||||
}
|
}
|
||||||
|
|
||||||
// Pop pops data from the start of the fifo
|
// Pop pops data from the start of the fifo
|
||||||
func (fifo *RedisByteFIFO) Pop() ([]byte, error) {
|
func (fifo *RedisByteFIFO) Pop(ctx context.Context) ([]byte, error) {
|
||||||
data, err := fifo.client.LPop(fifo.ctx, fifo.queueName).Bytes()
|
data, err := fifo.client.LPop(ctx, fifo.queueName).Bytes()
|
||||||
if err == nil || err == redis.Nil {
|
if err == nil || err == redis.Nil {
|
||||||
return data, nil
|
return data, nil
|
||||||
}
|
}
|
||||||
@ -122,8 +118,8 @@ func (fifo *RedisByteFIFO) Close() error {
|
|||||||
}
|
}
|
||||||
|
|
||||||
// Len returns the length of the fifo
|
// Len returns the length of the fifo
|
||||||
func (fifo *RedisByteFIFO) Len() int64 {
|
func (fifo *RedisByteFIFO) Len(ctx context.Context) int64 {
|
||||||
val, err := fifo.client.LLen(fifo.ctx, fifo.queueName).Result()
|
val, err := fifo.client.LLen(ctx, fifo.queueName).Result()
|
||||||
if err != nil {
|
if err != nil {
|
||||||
log.Error("Error whilst getting length of redis queue %s: Error: %v", fifo.queueName, err)
|
log.Error("Error whilst getting length of redis queue %s: Error: %v", fifo.queueName, err)
|
||||||
return -1
|
return -1
|
||||||
|
@ -38,7 +38,7 @@ type delayedStarter struct {
|
|||||||
}
|
}
|
||||||
|
|
||||||
// setInternal must be called with the lock locked.
|
// setInternal must be called with the lock locked.
|
||||||
func (q *delayedStarter) setInternal(atShutdown func(context.Context, func()), handle HandlerFunc, exemplar interface{}) error {
|
func (q *delayedStarter) setInternal(atShutdown func(func()), handle HandlerFunc, exemplar interface{}) error {
|
||||||
var ctx context.Context
|
var ctx context.Context
|
||||||
var cancel context.CancelFunc
|
var cancel context.CancelFunc
|
||||||
if q.timeout > 0 {
|
if q.timeout > 0 {
|
||||||
@ -49,9 +49,7 @@ func (q *delayedStarter) setInternal(atShutdown func(context.Context, func()), h
|
|||||||
|
|
||||||
defer cancel()
|
defer cancel()
|
||||||
// Ensure we also stop at shutdown
|
// Ensure we also stop at shutdown
|
||||||
atShutdown(ctx, func() {
|
atShutdown(cancel)
|
||||||
cancel()
|
|
||||||
})
|
|
||||||
|
|
||||||
i := 1
|
i := 1
|
||||||
for q.internal == nil {
|
for q.internal == nil {
|
||||||
@ -221,7 +219,7 @@ func (q *WrappedQueue) IsEmpty() bool {
|
|||||||
}
|
}
|
||||||
|
|
||||||
// Run starts to run the queue and attempts to create the internal queue
|
// Run starts to run the queue and attempts to create the internal queue
|
||||||
func (q *WrappedQueue) Run(atShutdown, atTerminate func(context.Context, func())) {
|
func (q *WrappedQueue) Run(atShutdown, atTerminate func(func())) {
|
||||||
log.Debug("WrappedQueue: %s Starting", q.name)
|
log.Debug("WrappedQueue: %s Starting", q.name)
|
||||||
q.lock.Lock()
|
q.lock.Lock()
|
||||||
if q.internal == nil {
|
if q.internal == nil {
|
||||||
|
@ -28,11 +28,15 @@ type ChannelUniqueQueueConfiguration ChannelQueueConfiguration
|
|||||||
// only guaranteed whilst the task is waiting in the queue.
|
// only guaranteed whilst the task is waiting in the queue.
|
||||||
type ChannelUniqueQueue struct {
|
type ChannelUniqueQueue struct {
|
||||||
*WorkerPool
|
*WorkerPool
|
||||||
lock sync.Mutex
|
lock sync.Mutex
|
||||||
table map[Data]bool
|
table map[Data]bool
|
||||||
exemplar interface{}
|
shutdownCtx context.Context
|
||||||
workers int
|
shutdownCtxCancel context.CancelFunc
|
||||||
name string
|
terminateCtx context.Context
|
||||||
|
terminateCtxCancel context.CancelFunc
|
||||||
|
exemplar interface{}
|
||||||
|
workers int
|
||||||
|
name string
|
||||||
}
|
}
|
||||||
|
|
||||||
// NewChannelUniqueQueue create a memory channel queue
|
// NewChannelUniqueQueue create a memory channel queue
|
||||||
@ -45,11 +49,19 @@ func NewChannelUniqueQueue(handle HandlerFunc, cfg, exemplar interface{}) (Queue
|
|||||||
if config.BatchLength == 0 {
|
if config.BatchLength == 0 {
|
||||||
config.BatchLength = 1
|
config.BatchLength = 1
|
||||||
}
|
}
|
||||||
|
|
||||||
|
terminateCtx, terminateCtxCancel := context.WithCancel(context.Background())
|
||||||
|
shutdownCtx, shutdownCtxCancel := context.WithCancel(terminateCtx)
|
||||||
|
|
||||||
queue := &ChannelUniqueQueue{
|
queue := &ChannelUniqueQueue{
|
||||||
table: map[Data]bool{},
|
table: map[Data]bool{},
|
||||||
exemplar: exemplar,
|
shutdownCtx: shutdownCtx,
|
||||||
workers: config.Workers,
|
shutdownCtxCancel: shutdownCtxCancel,
|
||||||
name: config.Name,
|
terminateCtx: terminateCtx,
|
||||||
|
terminateCtxCancel: terminateCtxCancel,
|
||||||
|
exemplar: exemplar,
|
||||||
|
workers: config.Workers,
|
||||||
|
name: config.Name,
|
||||||
}
|
}
|
||||||
queue.WorkerPool = NewWorkerPool(func(data ...Data) {
|
queue.WorkerPool = NewWorkerPool(func(data ...Data) {
|
||||||
for _, datum := range data {
|
for _, datum := range data {
|
||||||
@ -65,17 +77,11 @@ func NewChannelUniqueQueue(handle HandlerFunc, cfg, exemplar interface{}) (Queue
|
|||||||
}
|
}
|
||||||
|
|
||||||
// Run starts to run the queue
|
// Run starts to run the queue
|
||||||
func (q *ChannelUniqueQueue) Run(atShutdown, atTerminate func(context.Context, func())) {
|
func (q *ChannelUniqueQueue) Run(atShutdown, atTerminate func(func())) {
|
||||||
atShutdown(context.Background(), func() {
|
atShutdown(q.Shutdown)
|
||||||
log.Warn("ChannelUniqueQueue: %s is not shutdownable!", q.name)
|
atTerminate(q.Terminate)
|
||||||
})
|
|
||||||
atTerminate(context.Background(), func() {
|
|
||||||
log.Warn("ChannelUniqueQueue: %s is not terminatable!", q.name)
|
|
||||||
})
|
|
||||||
log.Debug("ChannelUniqueQueue: %s Starting", q.name)
|
log.Debug("ChannelUniqueQueue: %s Starting", q.name)
|
||||||
go func() {
|
_ = q.AddWorkers(q.workers, 0)
|
||||||
_ = q.AddWorkers(q.workers, 0)
|
|
||||||
}()
|
|
||||||
}
|
}
|
||||||
|
|
||||||
// Push will push data into the queue if the data is not already in the queue
|
// Push will push data into the queue if the data is not already in the queue
|
||||||
@ -122,6 +128,39 @@ func (q *ChannelUniqueQueue) Has(data Data) (bool, error) {
|
|||||||
return has, nil
|
return has, nil
|
||||||
}
|
}
|
||||||
|
|
||||||
|
// Shutdown processing from this queue
|
||||||
|
func (q *ChannelUniqueQueue) Shutdown() {
|
||||||
|
log.Trace("ChannelUniqueQueue: %s Shutting down", q.name)
|
||||||
|
select {
|
||||||
|
case <-q.shutdownCtx.Done():
|
||||||
|
return
|
||||||
|
default:
|
||||||
|
}
|
||||||
|
go func() {
|
||||||
|
log.Trace("ChannelUniqueQueue: %s Flushing", q.name)
|
||||||
|
if err := q.FlushWithContext(q.terminateCtx); err != nil {
|
||||||
|
log.Warn("ChannelUniqueQueue: %s Terminated before completed flushing", q.name)
|
||||||
|
return
|
||||||
|
}
|
||||||
|
log.Debug("ChannelUniqueQueue: %s Flushed", q.name)
|
||||||
|
}()
|
||||||
|
q.shutdownCtxCancel()
|
||||||
|
log.Debug("ChannelUniqueQueue: %s Shutdown", q.name)
|
||||||
|
}
|
||||||
|
|
||||||
|
// Terminate this queue and close the queue
|
||||||
|
func (q *ChannelUniqueQueue) Terminate() {
|
||||||
|
log.Trace("ChannelUniqueQueue: %s Terminating", q.name)
|
||||||
|
q.Shutdown()
|
||||||
|
select {
|
||||||
|
case <-q.terminateCtx.Done():
|
||||||
|
return
|
||||||
|
default:
|
||||||
|
}
|
||||||
|
q.terminateCtxCancel()
|
||||||
|
log.Debug("ChannelUniqueQueue: %s Terminated", q.name)
|
||||||
|
}
|
||||||
|
|
||||||
// Name returns the name of this queue
|
// Name returns the name of this queue
|
||||||
func (q *ChannelUniqueQueue) Name() string {
|
func (q *ChannelUniqueQueue) Name() string {
|
||||||
return q.name
|
return q.name
|
||||||
|
@ -5,6 +5,8 @@
|
|||||||
package queue
|
package queue
|
||||||
|
|
||||||
import (
|
import (
|
||||||
|
"context"
|
||||||
|
|
||||||
"code.gitea.io/gitea/modules/nosql"
|
"code.gitea.io/gitea/modules/nosql"
|
||||||
|
|
||||||
"gitea.com/lunny/levelqueue"
|
"gitea.com/lunny/levelqueue"
|
||||||
@ -41,6 +43,7 @@ func NewLevelUniqueQueue(handle HandlerFunc, cfg, exemplar interface{}) (Queue,
|
|||||||
if len(config.ConnectionString) == 0 {
|
if len(config.ConnectionString) == 0 {
|
||||||
config.ConnectionString = config.DataDir
|
config.ConnectionString = config.DataDir
|
||||||
}
|
}
|
||||||
|
config.WaitOnEmpty = true
|
||||||
|
|
||||||
byteFIFO, err := NewLevelUniqueQueueByteFIFO(config.ConnectionString, config.QueueName)
|
byteFIFO, err := NewLevelUniqueQueueByteFIFO(config.ConnectionString, config.QueueName)
|
||||||
if err != nil {
|
if err != nil {
|
||||||
@ -86,12 +89,12 @@ func NewLevelUniqueQueueByteFIFO(connection, prefix string) (*LevelUniqueQueueBy
|
|||||||
}
|
}
|
||||||
|
|
||||||
// PushFunc pushes data to the end of the fifo and calls the callback if it is added
|
// PushFunc pushes data to the end of the fifo and calls the callback if it is added
|
||||||
func (fifo *LevelUniqueQueueByteFIFO) PushFunc(data []byte, fn func() error) error {
|
func (fifo *LevelUniqueQueueByteFIFO) PushFunc(ctx context.Context, data []byte, fn func() error) error {
|
||||||
return fifo.internal.LPushFunc(data, fn)
|
return fifo.internal.LPushFunc(data, fn)
|
||||||
}
|
}
|
||||||
|
|
||||||
// Pop pops data from the start of the fifo
|
// Pop pops data from the start of the fifo
|
||||||
func (fifo *LevelUniqueQueueByteFIFO) Pop() ([]byte, error) {
|
func (fifo *LevelUniqueQueueByteFIFO) Pop(ctx context.Context) ([]byte, error) {
|
||||||
data, err := fifo.internal.RPop()
|
data, err := fifo.internal.RPop()
|
||||||
if err != nil && err != levelqueue.ErrNotFound {
|
if err != nil && err != levelqueue.ErrNotFound {
|
||||||
return nil, err
|
return nil, err
|
||||||
@ -100,12 +103,12 @@ func (fifo *LevelUniqueQueueByteFIFO) Pop() ([]byte, error) {
|
|||||||
}
|
}
|
||||||
|
|
||||||
// Len returns the length of the fifo
|
// Len returns the length of the fifo
|
||||||
func (fifo *LevelUniqueQueueByteFIFO) Len() int64 {
|
func (fifo *LevelUniqueQueueByteFIFO) Len(ctx context.Context) int64 {
|
||||||
return fifo.internal.Len()
|
return fifo.internal.Len()
|
||||||
}
|
}
|
||||||
|
|
||||||
// Has returns whether the fifo contains this data
|
// Has returns whether the fifo contains this data
|
||||||
func (fifo *LevelUniqueQueueByteFIFO) Has(data []byte) (bool, error) {
|
func (fifo *LevelUniqueQueueByteFIFO) Has(ctx context.Context, data []byte) (bool, error) {
|
||||||
return fifo.internal.Has(data)
|
return fifo.internal.Has(data)
|
||||||
}
|
}
|
||||||
|
|
||||||
|
@ -36,7 +36,7 @@ type PersistableChannelUniqueQueueConfiguration struct {
|
|||||||
// task cannot be processed twice or more at the same time. Uniqueness is
|
// task cannot be processed twice or more at the same time. Uniqueness is
|
||||||
// only guaranteed whilst the task is waiting in the queue.
|
// only guaranteed whilst the task is waiting in the queue.
|
||||||
type PersistableChannelUniqueQueue struct {
|
type PersistableChannelUniqueQueue struct {
|
||||||
*ChannelUniqueQueue
|
channelQueue *ChannelUniqueQueue
|
||||||
delayedStarter
|
delayedStarter
|
||||||
lock sync.Mutex
|
lock sync.Mutex
|
||||||
closed chan struct{}
|
closed chan struct{}
|
||||||
@ -85,8 +85,8 @@ func NewPersistableChannelUniqueQueue(handle HandlerFunc, cfg, exemplar interfac
|
|||||||
}
|
}
|
||||||
|
|
||||||
queue := &PersistableChannelUniqueQueue{
|
queue := &PersistableChannelUniqueQueue{
|
||||||
ChannelUniqueQueue: channelUniqueQueue.(*ChannelUniqueQueue),
|
channelQueue: channelUniqueQueue.(*ChannelUniqueQueue),
|
||||||
closed: make(chan struct{}),
|
closed: make(chan struct{}),
|
||||||
}
|
}
|
||||||
|
|
||||||
levelQueue, err := NewLevelUniqueQueue(func(data ...Data) {
|
levelQueue, err := NewLevelUniqueQueue(func(data ...Data) {
|
||||||
@ -138,14 +138,14 @@ func (q *PersistableChannelUniqueQueue) PushFunc(data Data, fn func() error) err
|
|||||||
case <-q.closed:
|
case <-q.closed:
|
||||||
return q.internal.(UniqueQueue).PushFunc(data, fn)
|
return q.internal.(UniqueQueue).PushFunc(data, fn)
|
||||||
default:
|
default:
|
||||||
return q.ChannelUniqueQueue.PushFunc(data, fn)
|
return q.channelQueue.PushFunc(data, fn)
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
// Has will test if the queue has the data
|
// Has will test if the queue has the data
|
||||||
func (q *PersistableChannelUniqueQueue) Has(data Data) (bool, error) {
|
func (q *PersistableChannelUniqueQueue) Has(data Data) (bool, error) {
|
||||||
// This is more difficult...
|
// This is more difficult...
|
||||||
has, err := q.ChannelUniqueQueue.Has(data)
|
has, err := q.channelQueue.Has(data)
|
||||||
if err != nil || has {
|
if err != nil || has {
|
||||||
return has, err
|
return has, err
|
||||||
}
|
}
|
||||||
@ -158,7 +158,7 @@ func (q *PersistableChannelUniqueQueue) Has(data Data) (bool, error) {
|
|||||||
}
|
}
|
||||||
|
|
||||||
// Run starts to run the queue
|
// Run starts to run the queue
|
||||||
func (q *PersistableChannelUniqueQueue) Run(atShutdown, atTerminate func(context.Context, func())) {
|
func (q *PersistableChannelUniqueQueue) Run(atShutdown, atTerminate func(func())) {
|
||||||
log.Debug("PersistableChannelUniqueQueue: %s Starting", q.delayedStarter.name)
|
log.Debug("PersistableChannelUniqueQueue: %s Starting", q.delayedStarter.name)
|
||||||
|
|
||||||
q.lock.Lock()
|
q.lock.Lock()
|
||||||
@ -170,7 +170,7 @@ func (q *PersistableChannelUniqueQueue) Run(atShutdown, atTerminate func(context
|
|||||||
log.Error("Unable push to channelled queue: %v", err)
|
log.Error("Unable push to channelled queue: %v", err)
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
}, q.exemplar)
|
}, q.channelQueue.exemplar)
|
||||||
q.lock.Unlock()
|
q.lock.Unlock()
|
||||||
if err != nil {
|
if err != nil {
|
||||||
log.Fatal("Unable to create internal queue for %s Error: %v", q.Name(), err)
|
log.Fatal("Unable to create internal queue for %s Error: %v", q.Name(), err)
|
||||||
@ -179,53 +179,73 @@ func (q *PersistableChannelUniqueQueue) Run(atShutdown, atTerminate func(context
|
|||||||
} else {
|
} else {
|
||||||
q.lock.Unlock()
|
q.lock.Unlock()
|
||||||
}
|
}
|
||||||
atShutdown(context.Background(), q.Shutdown)
|
atShutdown(q.Shutdown)
|
||||||
atTerminate(context.Background(), q.Terminate)
|
atTerminate(q.Terminate)
|
||||||
|
_ = q.channelQueue.AddWorkers(q.channelQueue.workers, 0)
|
||||||
|
|
||||||
// Just run the level queue - we shut it down later
|
if luq, ok := q.internal.(*LevelUniqueQueue); ok && luq.ByteFIFOUniqueQueue.byteFIFO.Len(luq.shutdownCtx) != 0 {
|
||||||
go q.internal.Run(func(_ context.Context, _ func()) {}, func(_ context.Context, _ func()) {})
|
// Just run the level queue - we shut it down once it's flushed
|
||||||
|
go q.internal.Run(func(_ func()) {}, func(_ func()) {})
|
||||||
|
go func() {
|
||||||
|
_ = q.internal.Flush(0)
|
||||||
|
log.Debug("LevelUniqueQueue: %s flushed so shutting down", q.internal.(*LevelQueue).Name())
|
||||||
|
q.internal.(*LevelUniqueQueue).Shutdown()
|
||||||
|
GetManager().Remove(q.internal.(*LevelUniqueQueue).qid)
|
||||||
|
}()
|
||||||
|
} else {
|
||||||
|
log.Debug("PersistableChannelUniqueQueue: %s Skipping running the empty level queue", q.delayedStarter.name)
|
||||||
|
q.internal.(*LevelUniqueQueue).Shutdown()
|
||||||
|
GetManager().Remove(q.internal.(*LevelUniqueQueue).qid)
|
||||||
|
}
|
||||||
|
|
||||||
go func() {
|
|
||||||
_ = q.ChannelUniqueQueue.AddWorkers(q.workers, 0)
|
|
||||||
}()
|
|
||||||
|
|
||||||
log.Trace("PersistableChannelUniqueQueue: %s Waiting til closed", q.delayedStarter.name)
|
|
||||||
<-q.closed
|
|
||||||
log.Trace("PersistableChannelUniqueQueue: %s Cancelling pools", q.delayedStarter.name)
|
|
||||||
q.internal.(*LevelUniqueQueue).cancel()
|
|
||||||
q.ChannelUniqueQueue.cancel()
|
|
||||||
log.Trace("PersistableChannelUniqueQueue: %s Waiting til done", q.delayedStarter.name)
|
|
||||||
q.ChannelUniqueQueue.Wait()
|
|
||||||
q.internal.(*LevelUniqueQueue).Wait()
|
|
||||||
// Redirect all remaining data in the chan to the internal channel
|
|
||||||
go func() {
|
|
||||||
log.Trace("PersistableChannelUniqueQueue: %s Redirecting remaining data", q.delayedStarter.name)
|
|
||||||
for data := range q.ChannelUniqueQueue.dataChan {
|
|
||||||
_ = q.internal.Push(data)
|
|
||||||
}
|
|
||||||
log.Trace("PersistableChannelUniqueQueue: %s Done Redirecting remaining data", q.delayedStarter.name)
|
|
||||||
}()
|
|
||||||
log.Trace("PersistableChannelUniqueQueue: %s Done main loop", q.delayedStarter.name)
|
|
||||||
}
|
}
|
||||||
|
|
||||||
// Flush flushes the queue
|
// Flush flushes the queue
|
||||||
func (q *PersistableChannelUniqueQueue) Flush(timeout time.Duration) error {
|
func (q *PersistableChannelUniqueQueue) Flush(timeout time.Duration) error {
|
||||||
return q.ChannelUniqueQueue.Flush(timeout)
|
return q.channelQueue.Flush(timeout)
|
||||||
|
}
|
||||||
|
|
||||||
|
// FlushWithContext flushes the queue
|
||||||
|
func (q *PersistableChannelUniqueQueue) FlushWithContext(ctx context.Context) error {
|
||||||
|
return q.channelQueue.FlushWithContext(ctx)
|
||||||
|
}
|
||||||
|
|
||||||
|
// IsEmpty checks if a queue is empty
|
||||||
|
func (q *PersistableChannelUniqueQueue) IsEmpty() bool {
|
||||||
|
return q.channelQueue.IsEmpty()
|
||||||
}
|
}
|
||||||
|
|
||||||
// Shutdown processing this queue
|
// Shutdown processing this queue
|
||||||
func (q *PersistableChannelUniqueQueue) Shutdown() {
|
func (q *PersistableChannelUniqueQueue) Shutdown() {
|
||||||
log.Trace("PersistableChannelUniqueQueue: %s Shutting down", q.delayedStarter.name)
|
log.Trace("PersistableChannelUniqueQueue: %s Shutting down", q.delayedStarter.name)
|
||||||
q.lock.Lock()
|
q.lock.Lock()
|
||||||
defer q.lock.Unlock()
|
|
||||||
select {
|
select {
|
||||||
case <-q.closed:
|
case <-q.closed:
|
||||||
|
q.lock.Unlock()
|
||||||
|
return
|
||||||
default:
|
default:
|
||||||
if q.internal != nil {
|
if q.internal != nil {
|
||||||
q.internal.(*LevelUniqueQueue).Shutdown()
|
q.internal.(*LevelUniqueQueue).Shutdown()
|
||||||
}
|
}
|
||||||
close(q.closed)
|
close(q.closed)
|
||||||
|
q.lock.Unlock()
|
||||||
}
|
}
|
||||||
|
|
||||||
|
log.Trace("PersistableChannelUniqueQueue: %s Cancelling pools", q.delayedStarter.name)
|
||||||
|
q.internal.(*LevelUniqueQueue).baseCtxCancel()
|
||||||
|
q.channelQueue.baseCtxCancel()
|
||||||
|
log.Trace("PersistableChannelUniqueQueue: %s Waiting til done", q.delayedStarter.name)
|
||||||
|
q.channelQueue.Wait()
|
||||||
|
q.internal.(*LevelUniqueQueue).Wait()
|
||||||
|
// Redirect all remaining data in the chan to the internal channel
|
||||||
|
go func() {
|
||||||
|
log.Trace("PersistableChannelUniqueQueue: %s Redirecting remaining data", q.delayedStarter.name)
|
||||||
|
for data := range q.channelQueue.dataChan {
|
||||||
|
_ = q.internal.Push(data)
|
||||||
|
}
|
||||||
|
log.Trace("PersistableChannelUniqueQueue: %s Done Redirecting remaining data", q.delayedStarter.name)
|
||||||
|
}()
|
||||||
|
|
||||||
log.Debug("PersistableChannelUniqueQueue: %s Shutdown", q.delayedStarter.name)
|
log.Debug("PersistableChannelUniqueQueue: %s Shutdown", q.delayedStarter.name)
|
||||||
}
|
}
|
||||||
|
|
||||||
|
@ -5,9 +5,8 @@
|
|||||||
package queue
|
package queue
|
||||||
|
|
||||||
import (
|
import (
|
||||||
"fmt"
|
"context"
|
||||||
|
|
||||||
"code.gitea.io/gitea/modules/graceful"
|
|
||||||
"github.com/go-redis/redis/v8"
|
"github.com/go-redis/redis/v8"
|
||||||
)
|
)
|
||||||
|
|
||||||
@ -51,8 +50,6 @@ func NewRedisUniqueQueue(handle HandlerFunc, cfg, exemplar interface{}) (Queue,
|
|||||||
return nil, err
|
return nil, err
|
||||||
}
|
}
|
||||||
|
|
||||||
byteFIFO.ctx = graceful.NewChannelContext(byteFIFOQueue.IsTerminated(), fmt.Errorf("queue has been terminated"))
|
|
||||||
|
|
||||||
queue := &RedisUniqueQueue{
|
queue := &RedisUniqueQueue{
|
||||||
ByteFIFOUniqueQueue: byteFIFOQueue,
|
ByteFIFOUniqueQueue: byteFIFOQueue,
|
||||||
}
|
}
|
||||||
@ -92,8 +89,8 @@ func NewRedisUniqueByteFIFO(config RedisUniqueByteFIFOConfiguration) (*RedisUniq
|
|||||||
}
|
}
|
||||||
|
|
||||||
// PushFunc pushes data to the end of the fifo and calls the callback if it is added
|
// PushFunc pushes data to the end of the fifo and calls the callback if it is added
|
||||||
func (fifo *RedisUniqueByteFIFO) PushFunc(data []byte, fn func() error) error {
|
func (fifo *RedisUniqueByteFIFO) PushFunc(ctx context.Context, data []byte, fn func() error) error {
|
||||||
added, err := fifo.client.SAdd(fifo.ctx, fifo.setName, data).Result()
|
added, err := fifo.client.SAdd(ctx, fifo.setName, data).Result()
|
||||||
if err != nil {
|
if err != nil {
|
||||||
return err
|
return err
|
||||||
}
|
}
|
||||||
@ -105,12 +102,12 @@ func (fifo *RedisUniqueByteFIFO) PushFunc(data []byte, fn func() error) error {
|
|||||||
return err
|
return err
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
return fifo.client.RPush(fifo.ctx, fifo.queueName, data).Err()
|
return fifo.client.RPush(ctx, fifo.queueName, data).Err()
|
||||||
}
|
}
|
||||||
|
|
||||||
// Pop pops data from the start of the fifo
|
// Pop pops data from the start of the fifo
|
||||||
func (fifo *RedisUniqueByteFIFO) Pop() ([]byte, error) {
|
func (fifo *RedisUniqueByteFIFO) Pop(ctx context.Context) ([]byte, error) {
|
||||||
data, err := fifo.client.LPop(fifo.ctx, fifo.queueName).Bytes()
|
data, err := fifo.client.LPop(ctx, fifo.queueName).Bytes()
|
||||||
if err != nil && err != redis.Nil {
|
if err != nil && err != redis.Nil {
|
||||||
return data, err
|
return data, err
|
||||||
}
|
}
|
||||||
@ -119,13 +116,13 @@ func (fifo *RedisUniqueByteFIFO) Pop() ([]byte, error) {
|
|||||||
return data, nil
|
return data, nil
|
||||||
}
|
}
|
||||||
|
|
||||||
err = fifo.client.SRem(fifo.ctx, fifo.setName, data).Err()
|
err = fifo.client.SRem(ctx, fifo.setName, data).Err()
|
||||||
return data, err
|
return data, err
|
||||||
}
|
}
|
||||||
|
|
||||||
// Has returns whether the fifo contains this data
|
// Has returns whether the fifo contains this data
|
||||||
func (fifo *RedisUniqueByteFIFO) Has(data []byte) (bool, error) {
|
func (fifo *RedisUniqueByteFIFO) Has(ctx context.Context, data []byte) (bool, error) {
|
||||||
return fifo.client.SIsMember(fifo.ctx, fifo.setName, data).Result()
|
return fifo.client.SIsMember(ctx, fifo.setName, data).Result()
|
||||||
}
|
}
|
||||||
|
|
||||||
func init() {
|
func init() {
|
||||||
|
@ -21,7 +21,7 @@ import (
|
|||||||
type WorkerPool struct {
|
type WorkerPool struct {
|
||||||
lock sync.Mutex
|
lock sync.Mutex
|
||||||
baseCtx context.Context
|
baseCtx context.Context
|
||||||
cancel context.CancelFunc
|
baseCtxCancel context.CancelFunc
|
||||||
cond *sync.Cond
|
cond *sync.Cond
|
||||||
qid int64
|
qid int64
|
||||||
maxNumberOfWorkers int
|
maxNumberOfWorkers int
|
||||||
@ -52,7 +52,7 @@ func NewWorkerPool(handle HandlerFunc, config WorkerPoolConfiguration) *WorkerPo
|
|||||||
dataChan := make(chan Data, config.QueueLength)
|
dataChan := make(chan Data, config.QueueLength)
|
||||||
pool := &WorkerPool{
|
pool := &WorkerPool{
|
||||||
baseCtx: ctx,
|
baseCtx: ctx,
|
||||||
cancel: cancel,
|
baseCtxCancel: cancel,
|
||||||
batchLength: config.BatchLength,
|
batchLength: config.BatchLength,
|
||||||
dataChan: dataChan,
|
dataChan: dataChan,
|
||||||
handle: handle,
|
handle: handle,
|
||||||
@ -83,7 +83,7 @@ func (p *WorkerPool) Push(data Data) {
|
|||||||
}
|
}
|
||||||
|
|
||||||
func (p *WorkerPool) zeroBoost() {
|
func (p *WorkerPool) zeroBoost() {
|
||||||
ctx, cancel := context.WithCancel(p.baseCtx)
|
ctx, cancel := context.WithTimeout(p.baseCtx, p.boostTimeout)
|
||||||
mq := GetManager().GetManagedQueue(p.qid)
|
mq := GetManager().GetManagedQueue(p.qid)
|
||||||
boost := p.boostWorkers
|
boost := p.boostWorkers
|
||||||
if (boost+p.numberOfWorkers) > p.maxNumberOfWorkers && p.maxNumberOfWorkers >= 0 {
|
if (boost+p.numberOfWorkers) > p.maxNumberOfWorkers && p.maxNumberOfWorkers >= 0 {
|
||||||
@ -94,26 +94,14 @@ func (p *WorkerPool) zeroBoost() {
|
|||||||
|
|
||||||
start := time.Now()
|
start := time.Now()
|
||||||
pid := mq.RegisterWorkers(boost, start, true, start.Add(p.boostTimeout), cancel, false)
|
pid := mq.RegisterWorkers(boost, start, true, start.Add(p.boostTimeout), cancel, false)
|
||||||
go func() {
|
cancel = func() {
|
||||||
select {
|
|
||||||
case <-ctx.Done():
|
|
||||||
case <-time.After(p.boostTimeout):
|
|
||||||
}
|
|
||||||
mq.RemoveWorkers(pid)
|
mq.RemoveWorkers(pid)
|
||||||
cancel()
|
}
|
||||||
}()
|
|
||||||
} else {
|
} else {
|
||||||
log.Warn("WorkerPool: %d has zero workers - adding %d temporary workers for %s", p.qid, p.boostWorkers, p.boostTimeout)
|
log.Warn("WorkerPool: %d has zero workers - adding %d temporary workers for %s", p.qid, p.boostWorkers, p.boostTimeout)
|
||||||
go func() {
|
|
||||||
select {
|
|
||||||
case <-ctx.Done():
|
|
||||||
case <-time.After(p.boostTimeout):
|
|
||||||
}
|
|
||||||
cancel()
|
|
||||||
}()
|
|
||||||
}
|
}
|
||||||
p.lock.Unlock()
|
p.lock.Unlock()
|
||||||
p.addWorkers(ctx, boost)
|
p.addWorkers(ctx, cancel, boost)
|
||||||
}
|
}
|
||||||
|
|
||||||
func (p *WorkerPool) pushBoost(data Data) {
|
func (p *WorkerPool) pushBoost(data Data) {
|
||||||
@ -140,7 +128,7 @@ func (p *WorkerPool) pushBoost(data Data) {
|
|||||||
return
|
return
|
||||||
}
|
}
|
||||||
p.blockTimeout *= 2
|
p.blockTimeout *= 2
|
||||||
ctx, cancel := context.WithCancel(p.baseCtx)
|
boostCtx, boostCtxCancel := context.WithCancel(p.baseCtx)
|
||||||
mq := GetManager().GetManagedQueue(p.qid)
|
mq := GetManager().GetManagedQueue(p.qid)
|
||||||
boost := p.boostWorkers
|
boost := p.boostWorkers
|
||||||
if (boost+p.numberOfWorkers) > p.maxNumberOfWorkers && p.maxNumberOfWorkers >= 0 {
|
if (boost+p.numberOfWorkers) > p.maxNumberOfWorkers && p.maxNumberOfWorkers >= 0 {
|
||||||
@ -150,24 +138,24 @@ func (p *WorkerPool) pushBoost(data Data) {
|
|||||||
log.Warn("WorkerPool: %d (for %s) Channel blocked for %v - adding %d temporary workers for %s, block timeout now %v", p.qid, mq.Name, ourTimeout, boost, p.boostTimeout, p.blockTimeout)
|
log.Warn("WorkerPool: %d (for %s) Channel blocked for %v - adding %d temporary workers for %s, block timeout now %v", p.qid, mq.Name, ourTimeout, boost, p.boostTimeout, p.blockTimeout)
|
||||||
|
|
||||||
start := time.Now()
|
start := time.Now()
|
||||||
pid := mq.RegisterWorkers(boost, start, true, start.Add(p.boostTimeout), cancel, false)
|
pid := mq.RegisterWorkers(boost, start, true, start.Add(p.boostTimeout), boostCtxCancel, false)
|
||||||
go func() {
|
go func() {
|
||||||
<-ctx.Done()
|
<-boostCtx.Done()
|
||||||
mq.RemoveWorkers(pid)
|
mq.RemoveWorkers(pid)
|
||||||
cancel()
|
boostCtxCancel()
|
||||||
}()
|
}()
|
||||||
} else {
|
} else {
|
||||||
log.Warn("WorkerPool: %d Channel blocked for %v - adding %d temporary workers for %s, block timeout now %v", p.qid, ourTimeout, p.boostWorkers, p.boostTimeout, p.blockTimeout)
|
log.Warn("WorkerPool: %d Channel blocked for %v - adding %d temporary workers for %s, block timeout now %v", p.qid, ourTimeout, p.boostWorkers, p.boostTimeout, p.blockTimeout)
|
||||||
}
|
}
|
||||||
go func() {
|
go func() {
|
||||||
<-time.After(p.boostTimeout)
|
<-time.After(p.boostTimeout)
|
||||||
cancel()
|
boostCtxCancel()
|
||||||
p.lock.Lock()
|
p.lock.Lock()
|
||||||
p.blockTimeout /= 2
|
p.blockTimeout /= 2
|
||||||
p.lock.Unlock()
|
p.lock.Unlock()
|
||||||
}()
|
}()
|
||||||
p.lock.Unlock()
|
p.lock.Unlock()
|
||||||
p.addWorkers(ctx, boost)
|
p.addWorkers(boostCtx, boostCtxCancel, boost)
|
||||||
p.dataChan <- data
|
p.dataChan <- data
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
@ -243,28 +231,25 @@ func (p *WorkerPool) commonRegisterWorkers(number int, timeout time.Duration, is
|
|||||||
mq := GetManager().GetManagedQueue(p.qid)
|
mq := GetManager().GetManagedQueue(p.qid)
|
||||||
if mq != nil {
|
if mq != nil {
|
||||||
pid := mq.RegisterWorkers(number, start, hasTimeout, end, cancel, isFlusher)
|
pid := mq.RegisterWorkers(number, start, hasTimeout, end, cancel, isFlusher)
|
||||||
go func() {
|
|
||||||
<-ctx.Done()
|
|
||||||
mq.RemoveWorkers(pid)
|
|
||||||
cancel()
|
|
||||||
}()
|
|
||||||
log.Trace("WorkerPool: %d (for %s) adding %d workers with group id: %d", p.qid, mq.Name, number, pid)
|
log.Trace("WorkerPool: %d (for %s) adding %d workers with group id: %d", p.qid, mq.Name, number, pid)
|
||||||
} else {
|
return ctx, func() {
|
||||||
log.Trace("WorkerPool: %d adding %d workers (no group id)", p.qid, number)
|
mq.RemoveWorkers(pid)
|
||||||
|
}
|
||||||
}
|
}
|
||||||
|
log.Trace("WorkerPool: %d adding %d workers (no group id)", p.qid, number)
|
||||||
|
|
||||||
return ctx, cancel
|
return ctx, cancel
|
||||||
}
|
}
|
||||||
|
|
||||||
// AddWorkers adds workers to the pool - this allows the number of workers to go above the limit
|
// AddWorkers adds workers to the pool - this allows the number of workers to go above the limit
|
||||||
func (p *WorkerPool) AddWorkers(number int, timeout time.Duration) context.CancelFunc {
|
func (p *WorkerPool) AddWorkers(number int, timeout time.Duration) context.CancelFunc {
|
||||||
ctx, cancel := p.commonRegisterWorkers(number, timeout, false)
|
ctx, cancel := p.commonRegisterWorkers(number, timeout, false)
|
||||||
p.addWorkers(ctx, number)
|
p.addWorkers(ctx, cancel, number)
|
||||||
return cancel
|
return cancel
|
||||||
}
|
}
|
||||||
|
|
||||||
// addWorkers adds workers to the pool
|
// addWorkers adds workers to the pool
|
||||||
func (p *WorkerPool) addWorkers(ctx context.Context, number int) {
|
func (p *WorkerPool) addWorkers(ctx context.Context, cancel context.CancelFunc, number int) {
|
||||||
for i := 0; i < number; i++ {
|
for i := 0; i < number; i++ {
|
||||||
p.lock.Lock()
|
p.lock.Lock()
|
||||||
if p.cond == nil {
|
if p.cond == nil {
|
||||||
@ -279,11 +264,13 @@ func (p *WorkerPool) addWorkers(ctx context.Context, number int) {
|
|||||||
p.numberOfWorkers--
|
p.numberOfWorkers--
|
||||||
if p.numberOfWorkers == 0 {
|
if p.numberOfWorkers == 0 {
|
||||||
p.cond.Broadcast()
|
p.cond.Broadcast()
|
||||||
|
cancel()
|
||||||
} else if p.numberOfWorkers < 0 {
|
} else if p.numberOfWorkers < 0 {
|
||||||
// numberOfWorkers can't go negative but...
|
// numberOfWorkers can't go negative but...
|
||||||
log.Warn("Number of Workers < 0 for QID %d - this shouldn't happen", p.qid)
|
log.Warn("Number of Workers < 0 for QID %d - this shouldn't happen", p.qid)
|
||||||
p.numberOfWorkers = 0
|
p.numberOfWorkers = 0
|
||||||
p.cond.Broadcast()
|
p.cond.Broadcast()
|
||||||
|
cancel()
|
||||||
}
|
}
|
||||||
p.lock.Unlock()
|
p.lock.Unlock()
|
||||||
}()
|
}()
|
||||||
|
@ -6,7 +6,6 @@
|
|||||||
package pull
|
package pull
|
||||||
|
|
||||||
import (
|
import (
|
||||||
"context"
|
|
||||||
"strconv"
|
"strconv"
|
||||||
"testing"
|
"testing"
|
||||||
"time"
|
"time"
|
||||||
@ -54,9 +53,9 @@ func TestPullRequest_AddToTaskQueue(t *testing.T) {
|
|||||||
assert.True(t, has)
|
assert.True(t, has)
|
||||||
assert.NoError(t, err)
|
assert.NoError(t, err)
|
||||||
|
|
||||||
prQueue.Run(func(_ context.Context, shutdown func()) {
|
prQueue.Run(func(shutdown func()) {
|
||||||
queueShutdown = append(queueShutdown, shutdown)
|
queueShutdown = append(queueShutdown, shutdown)
|
||||||
}, func(_ context.Context, terminate func()) {
|
}, func(terminate func()) {
|
||||||
queueTerminate = append(queueTerminate, terminate)
|
queueTerminate = append(queueTerminate, terminate)
|
||||||
})
|
})
|
||||||
|
|
||||||
|
Loading…
Reference in New Issue
Block a user