forgejo-federation/services/automerge/automerge.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

261 lines
7.7 KiB
Go

// Copyright 2021 Gitea. All rights reserved.
// SPDX-License-Identifier: MIT
package automerge
import (
"context"
"errors"
"fmt"
"strconv"
"strings"
"code.gitea.io/gitea/models/db"
issues_model "code.gitea.io/gitea/models/issues"
access_model "code.gitea.io/gitea/models/perm/access"
pull_model "code.gitea.io/gitea/models/pull"
repo_model "code.gitea.io/gitea/models/repo"
user_model "code.gitea.io/gitea/models/user"
"code.gitea.io/gitea/modules/git"
"code.gitea.io/gitea/modules/graceful"
"code.gitea.io/gitea/modules/log"
"code.gitea.io/gitea/modules/process"
"code.gitea.io/gitea/modules/queue"
pull_service "code.gitea.io/gitea/services/pull"
)
// prAutoMergeQueue represents a queue to handle update pull request tests
var prAutoMergeQueue *queue.WorkerPoolQueue[string]
// Init runs the task queue to that handles auto merges
func Init() error {
prAutoMergeQueue = queue.CreateUniqueQueue("pr_auto_merge", handler)
if prAutoMergeQueue == nil {
return fmt.Errorf("Unable to create pr_auto_merge Queue")
}
go graceful.GetManager().RunWithShutdownFns(prAutoMergeQueue.Run)
return nil
}
// handle passed PR IDs and test the PRs
func handler(items ...string) []string {
for _, s := range items {
var id int64
var sha string
if _, err := fmt.Sscanf(s, "%d_%s", &id, &sha); err != nil {
log.Error("could not parse data from pr_auto_merge queue (%v): %v", s, err)
continue
}
handlePull(id, sha)
}
return nil
}
func addToQueue(pr *issues_model.PullRequest, sha string) {
log.Trace("Adding pullID: %d to the pull requests patch checking queue with sha %s", pr.ID, sha)
if err := prAutoMergeQueue.Push(fmt.Sprintf("%d_%s", pr.ID, sha)); err != nil {
log.Error("Error adding pullID: %d to the pull requests patch checking queue %v", pr.ID, err)
}
}
// ScheduleAutoMerge if schedule is false and no error, pull can be merged directly
func ScheduleAutoMerge(ctx context.Context, doer *user_model.User, pull *issues_model.PullRequest, style repo_model.MergeStyle, message string) (scheduled bool, err error) {
err = db.WithTx(ctx, func(ctx context.Context) error {
lastCommitStatus, err := pull_service.GetPullRequestCommitStatusState(ctx, pull)
if err != nil {
return err
}
// we don't need to schedule
if lastCommitStatus.IsSuccess() {
return nil
}
if err := pull_model.ScheduleAutoMerge(ctx, doer, pull.ID, style, message); err != nil {
return err
}
scheduled = true
_, err = issues_model.CreateAutoMergeComment(ctx, issues_model.CommentTypePRScheduledToAutoMerge, pull, doer)
return err
})
return scheduled, err
}
// RemoveScheduledAutoMerge cancels a previously scheduled pull request
func RemoveScheduledAutoMerge(ctx context.Context, doer *user_model.User, pull *issues_model.PullRequest) error {
return db.WithTx(ctx, func(ctx context.Context) error {
if err := pull_model.DeleteScheduledAutoMerge(ctx, pull.ID); err != nil {
return err
}
_, err := issues_model.CreateAutoMergeComment(ctx, issues_model.CommentTypePRUnScheduledToAutoMerge, pull, doer)
return err
})
}
// MergeScheduledPullRequest merges a previously scheduled pull request when all checks succeeded
func MergeScheduledPullRequest(ctx context.Context, sha string, repo *repo_model.Repository) error {
pulls, err := getPullRequestsByHeadSHA(ctx, sha, repo, func(pr *issues_model.PullRequest) bool {
return !pr.HasMerged && pr.CanAutoMerge()
})
if err != nil {
return err
}
for _, pr := range pulls {
addToQueue(pr, sha)
}
return nil
}
func getPullRequestsByHeadSHA(ctx context.Context, sha string, repo *repo_model.Repository, filter func(*issues_model.PullRequest) bool) (map[int64]*issues_model.PullRequest, error) {
gitRepo, err := git.OpenRepository(ctx, repo.RepoPath())
if err != nil {
return nil, err
}
defer gitRepo.Close()
refs, err := gitRepo.GetRefsBySha(sha, "")
if err != nil {
return nil, err
}
pulls := make(map[int64]*issues_model.PullRequest)
for _, ref := range refs {
// Each pull branch starts with refs/pull/ we then go from there to find the index of the pr and then
// use that to get the pr.
if strings.HasPrefix(ref, git.PullPrefix) {
parts := strings.Split(ref[len(git.PullPrefix):], "/")
// e.g. 'refs/pull/1/head' would be []string{"1", "head"}
if len(parts) != 2 {
log.Error("getPullRequestsByHeadSHA found broken pull ref [%s] on repo [%-v]", ref, repo)
continue
}
prIndex, err := strconv.ParseInt(parts[0], 10, 64)
if err != nil {
log.Error("getPullRequestsByHeadSHA found broken pull ref [%s] on repo [%-v]", ref, repo)
continue
}
p, err := issues_model.GetPullRequestByIndex(ctx, repo.ID, prIndex)
if err != nil {
// If there is no pull request for this branch, we don't try to merge it.
if issues_model.IsErrPullRequestNotExist(err) {
continue
}
return nil, err
}
if filter(p) {
pulls[p.ID] = p
}
}
}
return pulls, nil
}
func handlePull(pullID int64, sha string) {
ctx, _, finished := process.GetManager().AddContext(graceful.GetManager().HammerContext(),
fmt.Sprintf("Handle AutoMerge of PR[%d] with sha[%s]", pullID, sha))
defer finished()
pr, err := issues_model.GetPullRequestByID(ctx, pullID)
if err != nil {
log.Error("GetPullRequestByID[%d]: %v", pullID, err)
return
}
// Check if there is a scheduled pr in the db
exists, scheduledPRM, err := pull_model.GetScheduledMergeByPullID(ctx, pr.ID)
if err != nil {
log.Error("%-v GetScheduledMergeByPullID: %v", pr, err)
return
}
if !exists {
return
}
// Get all checks for this pr
// We get the latest sha commit hash again to handle the case where the check of a previous push
// did not succeed or was not finished yet.
if err = pr.LoadHeadRepo(ctx); err != nil {
log.Error("%-v LoadHeadRepo: %v", pr, err)
return
}
headGitRepo, err := git.OpenRepository(ctx, pr.HeadRepo.RepoPath())
if err != nil {
log.Error("OpenRepository %-v: %v", pr.HeadRepo, err)
return
}
defer headGitRepo.Close()
headBranchExist := headGitRepo.IsBranchExist(pr.HeadBranch)
if pr.HeadRepo == nil || !headBranchExist {
log.Warn("Head branch of auto merge %-v does not exist [HeadRepoID: %d, Branch: %s]", pr, pr.HeadRepoID, pr.HeadBranch)
return
}
// Check if all checks succeeded
pass, err := pull_service.IsPullCommitStatusPass(ctx, pr)
if err != nil {
log.Error("%-v IsPullCommitStatusPass: %v", pr, err)
return
}
if !pass {
log.Info("Scheduled auto merge %-v has unsuccessful status checks", pr)
return
}
// Merge if all checks succeeded
doer, err := user_model.GetUserByID(ctx, scheduledPRM.DoerID)
if err != nil {
log.Error("Unable to get scheduled User[%d]: %v", scheduledPRM.DoerID, err)
return
}
perm, err := access_model.GetUserRepoPermission(ctx, pr.HeadRepo, doer)
if err != nil {
log.Error("GetUserRepoPermission %-v: %v", pr.HeadRepo, err)
return
}
if err := pull_service.CheckPullMergable(ctx, doer, &perm, pr, pull_service.MergeCheckTypeGeneral, false); err != nil {
if errors.Is(pull_service.ErrUserNotAllowedToMerge, err) {
log.Info("%-v was scheduled to automerge by an unauthorized user", pr)
return
}
log.Error("%-v CheckPullMergable: %v", pr, err)
return
}
var baseGitRepo *git.Repository
if pr.BaseRepoID == pr.HeadRepoID {
baseGitRepo = headGitRepo
} else {
if err = pr.LoadBaseRepo(ctx); err != nil {
log.Error("%-v LoadBaseRepo: %v", pr, err)
return
}
baseGitRepo, err = git.OpenRepository(ctx, pr.BaseRepo.RepoPath())
if err != nil {
log.Error("OpenRepository %-v: %v", pr.BaseRepo, err)
return
}
defer baseGitRepo.Close()
}
if err := pull_service.Merge(ctx, pr, doer, baseGitRepo, scheduledPRM.MergeStyle, "", scheduledPRM.Message, true); err != nil {
log.Error("pull_service.Merge: %v", err)
return
}
}