domi/forgejo-domi
1
0
Fork 0
forked from forgejo/forgejo
Code Pull requests Activity

Improve sync performance for pull-mirrors (#19125)

Browse source 
This addresses https://github.com/go-gitea/gitea/issues/18352

It aims to improve performance (and resource use) of the `SyncReleasesWithTags` operation for pull-mirrors.

For large repositories with many tags, `SyncReleasesWithTags` can be a costly operation (taking several minutes to complete). The reason is two-fold:
    
1. on sync, every upstream repo tag is compared (for changes) against existing local entries in the release table to ensure that they are up-to-date.
    
2. the procedure for getting _each tag_ involves a series of git operations    
    ```bash
     git show-ref --tags -- v8.2.4477
     git cat-file -t 29ab6ce9f36660cffaad3c8789e71162e5db5d2f
     git cat-file -p 29ab6ce9f36660cffaad3c8789e71162e5db5d2f
     git rev-list --count 29ab6ce9f36660cffaad3c8789e71162e5db5d2f
     ```    

     of which the `git rev-list --count` can be particularly heavy.
    
This PR optimizes performance for pull-mirrors. We utilize the fact that a pull-mirror is always identical to its upstream and rebuild the entire release table on every sync and use a batch `git for-each-ref .. refs/tags` call to retrieve all tags in one go.
    
For large mirror repos, with hundreds of annotated tags, this brings down the duration of the sync operation from several minutes to a few seconds. A few unscientific examples run on my local machine:

- https://github.com/spring-projects/spring-boot (223 tags)
  - before: `0m28,673s`
  - after: `0m2,244s`
- https://github.com/kubernetes/kubernetes (890 tags)
  - before: `8m00s`
  - after: `0m8,520s`
- https://github.com/vim/vim (13954 tags)
  - before: `14m20,383s`
  - after: `0m35,467s`

 

I added a `foreachref` package which contains a flexible way of specifying which reference fields are of interest (`git-for-each-ref(1)`) and to produce a parser for the expected output. These could be reused in other places where `for-each-ref` is used.  I'll add unit tests for those if the overall PR looks promising.
This commit is contained in:
Peter Gardfjäll 2022-03-31 14:30:40 +02:00 • committed by GitHub
parent b877504b03
commit e28cc79c92
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
7 changed files with 836 additions and 21 deletions
Download patch file Download diff file Expand all files Collapse all files

60
modules/repository/repo.go
View file

@ -150,6 +150,9 @@ func MigrateRepositoryGitData(ctx context.Context, u *user_model.User,
}
if !opts.Releases {
// note: this will greatly improve release (tag) sync
// for pull-mirrors with many tags
repo.IsMirror = opts.Mirror
if err = SyncReleasesWithTags(repo, gitRepo); err != nil {
log.Error("Failed to synchronize tags to releases for repository: %v", err)
}
@ -254,6 +257,14 @@ func CleanUpMigrateInfo(ctx context.Context, repo *repo_model.Repository) (*repo
// SyncReleasesWithTags synchronizes release table with repository tags
func SyncReleasesWithTags(repo *repo_model.Repository, gitRepo *git.Repository) error {
log.Debug("SyncReleasesWithTags: in Repo[%d:%s/%s]", repo.ID, repo.OwnerName, repo.Name)
// optimized procedure for pull-mirrors which saves a lot of time (in
// particular for repos with many tags).
if repo.IsMirror {
return pullMirrorReleaseSync(repo, gitRepo)
}
existingRelTags := make(map[string]struct{})
opts := models.FindReleasesOptions{
IncludeDrafts: true,
@ -450,3 +461,52 @@ func StoreMissingLfsObjectsInRepository(ctx context.Context, repo *repo_model.Re
return nil
}
// pullMirrorReleaseSync is a pull-mirror specific tag<->release table
// synchronization which overwrites all Releases from the repository tags. This
// can be relied on since a pull-mirror is always identical to its
// upstream. Hence, after each sync we want the pull-mirror release set to be
// identical to the upstream tag set. This is much more efficient for
// repositories like https://github.com/vim/vim (with over 13000 tags).
func pullMirrorReleaseSync(repo *repo_model.Repository, gitRepo *git.Repository) error {
log.Trace("pullMirrorReleaseSync: rebuilding releases for pull-mirror Repo[%d:%s/%s]", repo.ID, repo.OwnerName, repo.Name)
tags, numTags, err := gitRepo.GetTagInfos(0, 0)
if err != nil {
return fmt.Errorf("unable to GetTagInfos in pull-mirror Repo[%d:%s/%s]: %w", repo.ID, repo.OwnerName, repo.Name, err)
}
err = db.WithTx(func(ctx context.Context) error {
//
// clear out existing releases
//
if _, err := db.DeleteByBean(ctx, &models.Release{RepoID: repo.ID}); err != nil {
return fmt.Errorf("unable to clear releases for pull-mirror Repo[%d:%s/%s]: %w", repo.ID, repo.OwnerName, repo.Name, err)
}
//
// make release set identical to upstream tags
//
for _, tag := range tags {
release := models.Release{
RepoID: repo.ID,
TagName: tag.Name,
LowerTagName: strings.ToLower(tag.Name),
Sha1: tag.Object.String(),
// NOTE: ignored, since NumCommits are unused
// for pull-mirrors (only relevant when
// displaying releases, IsTag: false)
NumCommits: -1,
CreatedUnix: timeutil.TimeStamp(tag.Tagger.When.Unix()),
IsTag: true,
}
if err := db.Insert(ctx, release); err != nil {
return fmt.Errorf("unable insert tag %s for pull-mirror Repo[%d:%s/%s]: %w", tag.Name, repo.ID, repo.OwnerName, repo.Name, err)
}
}
return nil
})
if err != nil {
return fmt.Errorf("unable to rebuild release table for pull-mirror Repo[%d:%s/%s]: %w", repo.ID, repo.OwnerName, repo.Name, err)
}
log.Trace("pullMirrorReleaseSync: done rebuilding %d releases", numTags)
return nil
}