summaryrefslogtreecommitdiffstats
path: root/modules/queue/queue.go
diff options
context:
space:
mode:
authorwxiaoguang <wxiaoguang@gmail.com>2023-05-08 19:49:59 +0800
committerGitHub <noreply@github.com>2023-05-08 19:49:59 +0800
commit6f9c278559789066aa831c1df25b0d866103d02d (patch)
treee3a1880e0d4cf88916f9d1b65d82fbd4c41ea47f /modules/queue/queue.go
parentcb700aedd1e670fb47b8cf0cd67fb117a1ad88a2 (diff)
downloadgitea-6f9c278559789066aa831c1df25b0d866103d02d.tar.gz
gitea-6f9c278559789066aa831c1df25b0d866103d02d.zip
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)
Diffstat (limited to 'modules/queue/queue.go')
-rw-r--r--modules/queue/queue.go220
1 files changed, 25 insertions, 195 deletions
diff --git a/modules/queue/queue.go b/modules/queue/queue.go
index 22ee64f8e2..0ab8dd4ae4 100644
--- a/modules/queue/queue.go
+++ b/modules/queue/queue.go
@@ -1,201 +1,31 @@
-// Copyright 2019 The Gitea Authors. All rights reserved.
+// Copyright 2023 The Gitea Authors. All rights reserved.
// SPDX-License-Identifier: MIT
-package queue
-
-import (
- "context"
- "fmt"
- "time"
-)
-
-// ErrInvalidConfiguration is called when there is invalid configuration for a queue
-type ErrInvalidConfiguration struct {
- cfg interface{}
- err error
-}
-
-func (err ErrInvalidConfiguration) Error() string {
- if err.err != nil {
- return fmt.Sprintf("Invalid Configuration Argument: %v: Error: %v", err.cfg, err.err)
- }
- return fmt.Sprintf("Invalid Configuration Argument: %v", err.cfg)
-}
-
-// IsErrInvalidConfiguration checks if an error is an ErrInvalidConfiguration
-func IsErrInvalidConfiguration(err error) bool {
- _, ok := err.(ErrInvalidConfiguration)
- return ok
-}
-
-// Type is a type of Queue
-type Type string
-
-// Data defines an type of queuable data
-type Data interface{}
-
-// HandlerFunc is a function that takes a variable amount of data and processes it
-type HandlerFunc func(...Data) (unhandled []Data)
-
-// NewQueueFunc is a function that creates a queue
-type NewQueueFunc func(handler HandlerFunc, config, exemplar interface{}) (Queue, error)
-
-// Shutdownable represents a queue that can be shutdown
-type Shutdownable interface {
- Shutdown()
- Terminate()
-}
-
-// Named represents a queue with a name
-type Named interface {
- Name() string
-}
-
-// Queue defines an interface of a queue-like item
+// Package queue implements a specialized queue system for Gitea.
//
-// Queues will handle their own contents in the Run method
-type Queue interface {
- Flushable
- Run(atShutdown, atTerminate func(func()))
- Push(Data) error
-}
-
-// PushBackable queues can be pushed back to
-type PushBackable interface {
- // PushBack pushes data back to the top of the fifo
- PushBack(Data) error
-}
-
-// DummyQueueType is the type for the dummy queue
-const DummyQueueType Type = "dummy"
-
-// NewDummyQueue creates a new DummyQueue
-func NewDummyQueue(handler HandlerFunc, opts, exemplar interface{}) (Queue, error) {
- return &DummyQueue{}, nil
-}
-
-// DummyQueue represents an empty queue
-type DummyQueue struct{}
-
-// Run does nothing
-func (*DummyQueue) Run(_, _ func(func())) {}
-
-// Push fakes a push of data to the queue
-func (*DummyQueue) Push(Data) error {
- return nil
-}
-
-// PushFunc fakes a push of data to the queue with a function. The function is never run.
-func (*DummyQueue) PushFunc(Data, func() error) error {
- return nil
-}
-
-// Has always returns false as this queue never does anything
-func (*DummyQueue) Has(Data) (bool, error) {
- return false, nil
-}
-
-// Flush always returns nil
-func (*DummyQueue) Flush(time.Duration) error {
- return nil
-}
-
-// FlushWithContext always returns nil
-func (*DummyQueue) FlushWithContext(context.Context) error {
- return nil
-}
-
-// IsEmpty asserts that the queue is empty
-func (*DummyQueue) IsEmpty() bool {
- return true
-}
-
-// ImmediateType is the type to execute the function when push
-const ImmediateType Type = "immediate"
-
-// NewImmediate creates a new false queue to execute the function when push
-func NewImmediate(handler HandlerFunc, opts, exemplar interface{}) (Queue, error) {
- return &Immediate{
- handler: handler,
- }, nil
-}
-
-// Immediate represents an direct execution queue
-type Immediate struct {
- handler HandlerFunc
-}
-
-// Run does nothing
-func (*Immediate) Run(_, _ func(func())) {}
-
-// Push fakes a push of data to the queue
-func (q *Immediate) Push(data Data) error {
- return q.PushFunc(data, nil)
-}
-
-// PushFunc fakes a push of data to the queue with a function. The function is never run.
-func (q *Immediate) PushFunc(data Data, f func() error) error {
- if f != nil {
- if err := f(); err != nil {
- return err
- }
- }
- q.handler(data)
- return nil
-}
-
-// Has always returns false as this queue never does anything
-func (*Immediate) Has(Data) (bool, error) {
- return false, nil
-}
-
-// Flush always returns nil
-func (*Immediate) Flush(time.Duration) error {
- return nil
-}
-
-// FlushWithContext always returns nil
-func (*Immediate) FlushWithContext(context.Context) error {
- return nil
-}
-
-// IsEmpty asserts that the queue is empty
-func (*Immediate) IsEmpty() bool {
- return true
-}
-
-var queuesMap = map[Type]NewQueueFunc{
- DummyQueueType: NewDummyQueue,
- ImmediateType: NewImmediate,
-}
+// There are two major kinds of concepts:
+//
+// * The "base queue": channel, level, redis:
+// - They have the same abstraction, the same interface, and they are tested by the same testing code.
+// - The dummy(immediate) queue is special, it's not a real queue, it's only used as a no-op queue or a testing queue.
+//
+// * The WorkerPoolQueue: it uses the "base queue" to provide "worker pool" function.
+// - It calls the "handler" to process the data in the base queue.
+// - Its "Push" function doesn't block forever,
+// it will return an error if the queue is full after the timeout.
+//
+// A queue can be "simple" or "unique". A unique queue will try to avoid duplicate items.
+// Unique queue's "Has" function can be used to check whether an item is already in the queue,
+// although it's not 100% reliable due to there is no proper transaction support.
+// Simple queue's "Has" function always returns "has=false".
+//
+// The HandlerFuncT function is called by the WorkerPoolQueue to process the data in the base queue.
+// If the handler returns "unhandled" items, they will be re-queued to the base queue after a slight delay,
+// in case the item processor (eg: document indexer) is not available.
+package queue
-// RegisteredTypes provides the list of requested types of queues
-func RegisteredTypes() []Type {
- types := make([]Type, len(queuesMap))
- i := 0
- for key := range queuesMap {
- types[i] = key
- i++
- }
- return types
-}
+import "code.gitea.io/gitea/modules/util"
-// RegisteredTypesAsString provides the list of requested types of queues
-func RegisteredTypesAsString() []string {
- types := make([]string, len(queuesMap))
- i := 0
- for key := range queuesMap {
- types[i] = string(key)
- i++
- }
- return types
-}
+type HandlerFuncT[T any] func(...T) (unhandled []T)
-// NewQueue takes a queue Type, HandlerFunc, some options and possibly an exemplar and returns a Queue or an error
-func NewQueue(queueType Type, handlerFunc HandlerFunc, opts, exemplar interface{}) (Queue, error) {
- newFn, ok := queuesMap[queueType]
- if !ok {
- return nil, fmt.Errorf("unsupported queue type: %v", queueType)
- }
- return newFn(handlerFunc, opts, exemplar)
-}
+var ErrAlreadyInQueue = util.NewAlreadyExistErrorf("already in queue")