Skip to content

Fix Go worker heartbeat cancellation handling #68535

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
jx2lee wants to merge 1 commit into
base: main
Choose a base branch
from
+35 −1
Open

Fix Go worker heartbeat cancellation handling #68535

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion go-sdk/pkg/worker/runner.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -327,7 +327,7 @@ func (w *worker) ExecuteTaskWorkload(ctx context.Context, workload api.ExecuteTa
var finalState api.TerminalTIState
body := &api.TIUpdateStatePayload{}

if taskContext.Err() == ErrTaskCancelledAfterFailedHeartbeat {
if errors.Is(context.Cause(taskContext), ErrTaskCancelledAfterFailedHeartbeat) {
// We've already logged when we failed to heartbeat, don't do it again
finalState = api.TerminalTIStateFailed
body.FromTITerminalStatePayload(api.TITerminalStatePayload{
Expand Down
34 changes: 34 additions & 0 deletions go-sdk/pkg/worker/runner_test.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -21,6 +21,7 @@ import (
"context"
"fmt"
"log/slog"
"net/http"
"sync/atomic"
"testing"
"time"
Expand All @@ -31,6 +32,7 @@ import (
"github.com/spf13/viper"
"github.com/stretchr/testify/mock"
"github.com/stretchr/testify/suite"
"resty.dev/v3"

"github.com/apache/airflow/go-sdk/bundle/bundlev1"
"github.com/apache/airflow/go-sdk/pkg/api"
Expand Down Expand Up @@ -218,6 +220,38 @@ func (s *WorkerSuite) TestTaskHeartbeatsWhileRunning() {
True(count <= 11 && count >= 9, fmt.Sprintf("Call count of %d was not within the margin of error of 10+/-1", count))
}

func (s *WorkerSuite) TestTaskHeartbeatConflictStopsTask() {
id := uuid.New().String()
testWorkload := newTestWorkLoad(id, id[:8])

s.registry.AddDag(testWorkload.TI.DagId).
AddTaskWithName(testWorkload.TI.TaskId, func(ctx context.Context) error {
select {
case <-ctx.Done():
return nil
case <-time.After(2 * time.Second):
return fmt.Errorf("task context was not cancelled")
}
})

s.ExpectTaskRun(id)
s.ExpectTaskState(id, api.TerminalTIStateFailed)
s.ti.EXPECT().
Heartbeat(mock.Anything, uuid.MustParse(id), mock.Anything).
Return(&api.GeneralHTTPError{
Response: &resty.Response{
RawResponse: &http.Response{
Status: "409 Conflict",
StatusCode: http.StatusConflict,
},
},
})
s.client.EXPECT().TaskInstances().Return(s.ti)

err := s.worker.ExecuteTaskWorkload(context.Background(), testWorkload)
s.NoError(err)
}

func (s *WorkerSuite) TestTaskHeartbeatErrorStopsTaskAndLogs() {
s.T().Skip("TODO: Not implemented yet")
}
Expand Down