Files
Gitea/models/actions/run_attempt.go
T
Zettat123 c186cc4b8d fix(actions): let a rerun of selected jobs read the previous attempt's artifacts (#38857)
Fixes #38773

## Background

Artifacts became attempt-scoped in #37119, and the runner-facing
artifact APIs filter strictly by the attempt of the running job. "Re-run
failed jobs" creates a new attempt whose passed-through jobs never
upload their artifacts again, so a re-run job that downloads one of them
fails with "artifact not found".

## Fix

The read paths (v3 and v4 list and download) now resolve artifacts
across the running job's attempt plus the attempts it inherits from, and
an inherited artifact is shadowed by a same-named one from a newer
attempt.

## Note

GitHub's documentation does not document these behaviors. The
conclusions below are based on manual testing, so consistency with
GitHub cannot be guaranteed.

- In a "partial re-run", a job can download artifacts uploaded by an
earlier attempt, every attempt keeps its own copy of a name, and a
lookup by name resolves to the newest one.
- A full "Re-run all jobs" never downloads artifacts from earlier
attempts.

---------

Co-authored-by: bircni <bircni@icloud.com>
2026-08-12 19:55:26 +00:00

194 lines
6.2 KiB
Go

// Copyright 2026 The Gitea Authors. All rights reserved.
// SPDX-License-Identifier: MIT
package actions
import (
"context"
"fmt"
"slices"
"time"
"gitea.dev/models/db"
user_model "gitea.dev/models/user"
"gitea.dev/modules/container"
"gitea.dev/modules/log"
"gitea.dev/modules/timeutil"
"gitea.dev/modules/util"
)
// ActionRunAttempt represents a single execution attempt of an ActionRun.
type ActionRunAttempt struct {
ID int64
RepoID int64 `xorm:"index(repo_concurrency_status)"`
RunID int64 `xorm:"UNIQUE(run_attempt)"`
Run *ActionRun `xorm:"-"`
Attempt int64 `xorm:"UNIQUE(run_attempt)"`
TriggerUserID int64
TriggerUser *user_model.User `xorm:"-"`
ConcurrencyGroup string `xorm:"index(repo_concurrency_status) NOT NULL DEFAULT ''"`
ConcurrencyCancel bool `xorm:"NOT NULL DEFAULT FALSE"`
Status Status `xorm:"index(repo_concurrency_status)"`
Started timeutil.TimeStamp
Stopped timeutil.TimeStamp
Created timeutil.TimeStamp `xorm:"created"`
Updated timeutil.TimeStamp `xorm:"updated"`
}
func (*ActionRunAttempt) TableName() string {
return "action_run_attempt"
}
func init() {
db.RegisterModel(new(ActionRunAttempt))
}
func (attempt *ActionRunAttempt) Duration() time.Duration {
return calculateDuration(attempt.Started, attempt.Stopped, attempt.Status, attempt.Updated)
}
func (attempt *ActionRunAttempt) LoadAttributes(ctx context.Context) (err error) {
if attempt.Run == nil {
run, err := GetRunByRepoAndID(ctx, attempt.RepoID, attempt.RunID)
if err != nil {
return err
}
if err := run.LoadAttributes(ctx); err != nil {
return err
}
attempt.Run = run
}
return attempt.LoadTriggerUser(ctx)
}
// LoadTriggerUser loads the attempt's trigger user if not already loaded.
func (attempt *ActionRunAttempt) LoadTriggerUser(ctx context.Context) (err error) {
if attempt.TriggerUser != nil {
return nil
}
attempt.TriggerUserID, attempt.TriggerUser, err = user_model.GetPossibleUserByID(ctx, attempt.TriggerUserID)
return err
}
func GetRunAttemptByRepoAndID(ctx context.Context, repoID, attemptID int64) (*ActionRunAttempt, error) {
var attempt ActionRunAttempt
has, err := db.GetEngine(ctx).Where("repo_id=? AND id=?", repoID, attemptID).Get(&attempt)
if err != nil {
return nil, err
} else if !has {
return nil, fmt.Errorf("run attempt %d in repo %d: %w", attemptID, repoID, util.ErrNotExist)
}
return &attempt, nil
}
func GetRunAttemptByRunIDAndAttemptNum(ctx context.Context, runID, attemptNum int64) (*ActionRunAttempt, error) {
var attempt ActionRunAttempt
has, err := db.GetEngine(ctx).Where("run_id=? AND attempt=?", runID, attemptNum).Get(&attempt)
if err != nil {
return nil, err
} else if !has {
return nil, fmt.Errorf("run attempt %d for run %d: %w", attemptNum, runID, util.ErrNotExist)
}
return &attempt, nil
}
// GetArtifactAttemptIDs returns the IDs of the attempts whose artifacts the job may read, newest first,
// always including the job's own attempt.
// An attempt that re-ran only some of the run's jobs keeps the artifacts of the attempt it re-ran from,
// because the jobs it passed through never upload them again; a rerun of the whole run starts over.
func GetArtifactAttemptIDs(ctx context.Context, job *ActionRunJob) ([]int64, error) {
if job.Attempt <= 1 || job.RunAttemptID == 0 {
return []int64{job.RunAttemptID}, nil
}
attempts, err := ListRunAttemptsByRunID(ctx, job.RunID)
if err != nil {
return nil, err
}
// a newer attempt is never readable, and attempt 1 has nothing older to continue into
candidateIDs := container.FilterSlice(attempts, func(a *ActionRunAttempt) (int64, bool) {
return a.ID, a.Attempt > 1 && a.Attempt <= job.Attempt
})
passThroughAttemptIDs, err := findPassThroughAttemptIDs(ctx, candidateIDs)
if err != nil {
return nil, err
}
ids := make([]int64, 0, len(attempts))
for _, attempt := range attempts {
if attempt.Attempt > job.Attempt {
continue
}
ids = append(ids, attempt.ID)
if !slices.Contains(passThroughAttemptIDs, attempt.ID) {
// stops at the first attempt that passed no job through
break
}
}
return ids, nil
}
// findPassThroughAttemptIDs narrows the given attempts to those that were a rerun of selected jobs:
// only such a rerun clones jobs carrying a source task.
// TODO: best-effort. Needs a better way to distinguish between "partial re-run" and "full re-run".
func findPassThroughAttemptIDs(ctx context.Context, attemptIDs []int64) ([]int64, error) {
passThroughAttemptIDs := make([]int64, 0, len(attemptIDs))
return passThroughAttemptIDs, db.GetEngine(ctx).
Table("action_run_job").
Cols("run_attempt_id").
In("run_attempt_id", attemptIDs).
Where("source_task_id <> 0").
Distinct("run_attempt_id").
Find(&passThroughAttemptIDs)
}
// FindConcurrentRunAttempts returns attempts in the given concurrency group and status set.
// Results are unordered; callers must not depend on any particular row order.
func FindConcurrentRunAttempts(ctx context.Context, repoID int64, concurrencyGroup string, statuses []Status) ([]*ActionRunAttempt, error) {
attempts := make([]*ActionRunAttempt, 0)
sess := db.GetEngine(ctx).Where("repo_id=? AND concurrency_group=?", repoID, concurrencyGroup)
if len(statuses) > 0 {
sess = sess.In("status", statuses)
}
return attempts, sess.Find(&attempts)
}
func UpdateRunAttempt(ctx context.Context, attempt *ActionRunAttempt, cols ...string) error {
if slices.Contains(cols, "status") && attempt.Started.IsZero() && attempt.Status.IsRunning() {
attempt.Started = timeutil.TimeStampNow()
cols = append(cols, "started")
}
sess := db.GetEngine(ctx).ID(attempt.ID)
if len(cols) > 0 {
sess.Cols(cols...)
}
if _, err := sess.Update(attempt); err != nil {
return err
}
// Only status/timing changes on an attempt need to update the latest run.
if len(cols) > 0 && !slices.Contains(cols, "status") && !slices.Contains(cols, "started") && !slices.Contains(cols, "stopped") {
return nil
}
run, err := GetRunByRepoAndID(ctx, attempt.RepoID, attempt.RunID)
if err != nil {
return err
}
if run.LatestAttemptID != attempt.ID {
log.Warn("run %d cannot be updated by an old attempt %d", run.LatestAttemptID, attempt.ID)
return nil
}
run.Status = attempt.Status
run.Started = attempt.Started
run.Stopped = attempt.Stopped
return UpdateRun(ctx, run, "status", "started", "stopped")
}