1
0
Fork 0
mirror of https://codeberg.org/forgejo/forgejo.git synced 2024-12-28 13:49:13 -05:00
forgejo/modules/queue/workerqueue.go
wxiaoguang 6f9c278559
Rewrite queue (#24505)
# ⚠️ Breaking

Many deprecated queue config options are removed (actually, they should
have been removed in 1.18/1.19).

If you see the fatal message when starting Gitea: "Please update your
app.ini to remove deprecated config options", please follow the error
messages to remove these options from your app.ini.

Example:

```
2023/05/06 19:39:22 [E] Removed queue option: `[indexer].ISSUE_INDEXER_QUEUE_TYPE`. Use new options in `[queue.issue_indexer]`
2023/05/06 19:39:22 [E] Removed queue option: `[indexer].UPDATE_BUFFER_LEN`. Use new options in `[queue.issue_indexer]`
2023/05/06 19:39:22 [F] Please update your app.ini to remove deprecated config options
```

Many options in `[queue]` are are dropped, including:
`WRAP_IF_NECESSARY`, `MAX_ATTEMPTS`, `TIMEOUT`, `WORKERS`,
`BLOCK_TIMEOUT`, `BOOST_TIMEOUT`, `BOOST_WORKERS`, they can be removed
from app.ini.

# The problem

The old queue package has some legacy problems:

* complexity: I doubt few people could tell how it works.
* maintainability: Too many channels and mutex/cond are mixed together,
too many different structs/interfaces depends each other.
* stability: due to the complexity & maintainability, sometimes there
are strange bugs and difficult to debug, and some code doesn't have test
(indeed some code is difficult to test because a lot of things are mixed
together).
* general applicability: although it is called "queue", its behavior is
not a well-known queue.
* scalability: it doesn't seem easy to make it work with a cluster
without breaking its behaviors.

It came from some very old code to "avoid breaking", however, its
technical debt is too heavy now. It's a good time to introduce a better
"queue" package.

# The new queue package

It keeps using old config and concept as much as possible.

* It only contains two major kinds of concepts:
    * The "base queue": channel, levelqueue, redis
* They have the same abstraction, the same interface, and they are
tested by the same testing code.
* The "WokerPoolQueue", it uses the "base queue" to provide "worker
pool" function, calls the "handler" to process the data in the base
queue.
* The new code doesn't do "PushBack"
* Think about a queue with many workers, the "PushBack" can't guarantee
the order for re-queued unhandled items, so in new code it just does
"normal push"
* The new code doesn't do "pause/resume"
* The "pause/resume" was designed to handle some handler's failure: eg:
document indexer (elasticsearch) is down
* If a queue is paused for long time, either the producers blocks or the
new items are dropped.
* The new code doesn't do such "pause/resume" trick, it's not a common
queue's behavior and it doesn't help much.
* If there are unhandled items, the "push" function just blocks for a
few seconds and then re-queue them and retry.
* The new code doesn't do "worker booster"
* Gitea's queue's handlers are light functions, the cost is only the
go-routine, so it doesn't make sense to "boost" them.
* The new code only use "max worker number" to limit the concurrent
workers.
* The new "Push" never blocks forever
* Instead of creating more and more blocking goroutines, return an error
is more friendly to the server and to the end user.

There are more details in code comments: eg: the "Flush" problem, the
strange "code.index" hanging problem, the "immediate" queue problem.

Almost ready for review.

TODO:

* [x] add some necessary comments during review
* [x] add some more tests if necessary
* [x] update documents and config options
* [x] test max worker / active worker
* [x] re-run the CI tasks to see whether any test is flaky
* [x] improve the `handleOldLengthConfiguration` to provide more
friendly messages
* [x] fine tune default config values (eg: length?)

## Code coverage:

![image](https://user-images.githubusercontent.com/2114189/236620635-55576955-f95d-4810-b12f-879026a3afdf.png)
2023-05-08 19:49:59 +08:00

241 lines
6.6 KiB
Go

// Copyright 2023 The Gitea Authors. All rights reserved.
// SPDX-License-Identifier: MIT
package queue
import (
"context"
"fmt"
"sync"
"sync/atomic"
"time"
"code.gitea.io/gitea/modules/graceful"
"code.gitea.io/gitea/modules/json"
"code.gitea.io/gitea/modules/log"
"code.gitea.io/gitea/modules/setting"
)
// WorkerPoolQueue is a queue that uses a pool of workers to process items
// It can use different underlying (base) queue types
type WorkerPoolQueue[T any] struct {
ctxRun context.Context
ctxRunCancel context.CancelFunc
ctxShutdown atomic.Pointer[context.Context]
shutdownDone chan struct{}
origHandler HandlerFuncT[T]
safeHandler HandlerFuncT[T]
baseQueueType string
baseConfig *BaseConfig
baseQueue baseQueue
batchChan chan []T
flushChan chan flushType
batchLength int
workerNum int
workerMaxNum int
workerActiveNum int
workerNumMu sync.Mutex
}
type flushType chan struct{}
var _ ManagedWorkerPoolQueue = (*WorkerPoolQueue[any])(nil)
func (q *WorkerPoolQueue[T]) GetName() string {
return q.baseConfig.ManagedName
}
func (q *WorkerPoolQueue[T]) GetType() string {
return q.baseQueueType
}
func (q *WorkerPoolQueue[T]) GetItemTypeName() string {
var t T
return fmt.Sprintf("%T", t)
}
func (q *WorkerPoolQueue[T]) GetWorkerNumber() int {
q.workerNumMu.Lock()
defer q.workerNumMu.Unlock()
return q.workerNum
}
func (q *WorkerPoolQueue[T]) GetWorkerActiveNumber() int {
q.workerNumMu.Lock()
defer q.workerNumMu.Unlock()
return q.workerActiveNum
}
func (q *WorkerPoolQueue[T]) GetWorkerMaxNumber() int {
q.workerNumMu.Lock()
defer q.workerNumMu.Unlock()
return q.workerMaxNum
}
func (q *WorkerPoolQueue[T]) SetWorkerMaxNumber(num int) {
q.workerNumMu.Lock()
defer q.workerNumMu.Unlock()
q.workerMaxNum = num
}
func (q *WorkerPoolQueue[T]) GetQueueItemNumber() int {
cnt, err := q.baseQueue.Len(q.ctxRun)
if err != nil {
log.Error("Failed to get number of items in queue %q: %v", q.GetName(), err)
}
return cnt
}
func (q *WorkerPoolQueue[T]) FlushWithContext(ctx context.Context, timeout time.Duration) (err error) {
if q.isBaseQueueDummy() {
return
}
log.Debug("Try to flush queue %q with timeout %v", q.GetName(), timeout)
defer log.Debug("Finish flushing queue %q, err: %v", q.GetName(), err)
var after <-chan time.Time
after = infiniteTimerC
if timeout > 0 {
after = time.After(timeout)
}
c := make(flushType)
// send flush request
// if it blocks, it means that there is a flush in progress or the queue hasn't been started yet
select {
case q.flushChan <- c:
case <-ctx.Done():
return ctx.Err()
case <-q.ctxRun.Done():
return q.ctxRun.Err()
case <-after:
return context.DeadlineExceeded
}
// wait for flush to finish
select {
case <-c:
return nil
case <-ctx.Done():
return ctx.Err()
case <-q.ctxRun.Done():
return q.ctxRun.Err()
case <-after:
return context.DeadlineExceeded
}
}
func (q *WorkerPoolQueue[T]) marshal(data T) []byte {
bs, err := json.Marshal(data)
if err != nil {
log.Error("Failed to marshal item for queue %q: %v", q.GetName(), err)
return nil
}
return bs
}
func (q *WorkerPoolQueue[T]) unmarshal(data []byte) (t T, ok bool) {
if err := json.Unmarshal(data, &t); err != nil {
log.Error("Failed to unmarshal item from queue %q: %v", q.GetName(), err)
return t, false
}
return t, true
}
func (q *WorkerPoolQueue[T]) isBaseQueueDummy() bool {
_, isDummy := q.baseQueue.(*baseDummy)
return isDummy
}
// Push adds an item to the queue, it may block for a while and then returns an error if the queue is full
func (q *WorkerPoolQueue[T]) Push(data T) error {
if q.isBaseQueueDummy() && q.safeHandler != nil {
// FIXME: the "immediate" queue is only for testing, but it really causes problems because its behavior is different from a real queue.
// Even if tests pass, it doesn't mean that there is no bug in code.
if data, ok := q.unmarshal(q.marshal(data)); ok {
q.safeHandler(data)
}
}
return q.baseQueue.PushItem(q.ctxRun, q.marshal(data))
}
// Has only works for unique queues. Keep in mind that this check may not be reliable (due to lacking of proper transaction support)
// There could be a small chance that duplicate items appear in the queue
func (q *WorkerPoolQueue[T]) Has(data T) (bool, error) {
return q.baseQueue.HasItem(q.ctxRun, q.marshal(data))
}
func (q *WorkerPoolQueue[T]) Run(atShutdown, atTerminate func(func())) {
atShutdown(func() {
// in case some queue handlers are slow or have hanging bugs, at most wait for a short time
q.ShutdownWait(1 * time.Second)
})
q.doRun()
}
// ShutdownWait shuts down the queue, waits for all workers to finish their jobs, and pushes the unhandled items back to the base queue
// It waits for all workers (handlers) to finish their jobs, in case some buggy handlers would hang forever, a reasonable timeout is needed
func (q *WorkerPoolQueue[T]) ShutdownWait(timeout time.Duration) {
shutdownCtx, shutdownCtxCancel := context.WithTimeout(context.Background(), timeout)
defer shutdownCtxCancel()
if q.ctxShutdown.CompareAndSwap(nil, &shutdownCtx) {
q.ctxRunCancel()
}
<-q.shutdownDone
}
func getNewQueueFn(t string) (string, func(cfg *BaseConfig, unique bool) (baseQueue, error)) {
switch t {
case "dummy", "immediate":
return t, newBaseDummy
case "channel":
return t, newBaseChannelGeneric
case "redis":
return t, newBaseRedisGeneric
default: // level(leveldb,levelqueue,persistable-channel)
return "level", newBaseLevelQueueGeneric
}
}
func NewWorkerPoolQueueBySetting[T any](name string, queueSetting setting.QueueSettings, handler HandlerFuncT[T], unique bool) (*WorkerPoolQueue[T], error) {
if handler == nil {
log.Debug("Use dummy queue for %q because handler is nil and caller doesn't want to process the queue items", name)
queueSetting.Type = "dummy"
}
var w WorkerPoolQueue[T]
var err error
queueType, newQueueFn := getNewQueueFn(queueSetting.Type)
w.baseQueueType = queueType
w.baseConfig = toBaseConfig(name, queueSetting)
w.baseQueue, err = newQueueFn(w.baseConfig, unique)
if err != nil {
return nil, err
}
log.Trace("Created queue %q of type %q", name, queueType)
w.ctxRun, w.ctxRunCancel = context.WithCancel(graceful.GetManager().ShutdownContext())
w.batchChan = make(chan []T)
w.flushChan = make(chan flushType)
w.shutdownDone = make(chan struct{})
w.workerMaxNum = queueSetting.MaxWorkers
w.batchLength = queueSetting.BatchLength
w.origHandler = handler
w.safeHandler = func(t ...T) (unhandled []T) {
defer func() {
err := recover()
if err != nil {
log.Error("Recovered from panic in queue %q handler: %v\n%s", name, err, log.Stack(2))
}
}()
return w.origHandler(t...)
}
return &w, nil
}