temporalio / temporalio/temporal
WorkflowTaskTimedOut if submiting a large number of activiteis within one workflow
Nobody has claimed this yet.
- Dominant language
- Go
- Stars
- 23.2k
- Forks
- 1.9k
- Avg merge
- 2d 8h
- Merged PRs (30d)
- 228
Description
hi, I apologize for using an issue to seek help.
Expected Behavior
In my workflow worker, firstly I start a timer to wait for the biz data to be ready. Then 1,000 activies are submitted, i expected that all these activities can be scheduled and run asynchronously after the submission. Then I wait for the results of the 1,000 Future object like a wait group model.
This is a simplified version of my code:
func (w *Worker) Start(wfCtx workflow.Context, req *Req) (result string, err error) {
if err := workflow.Sleep(wfCtx, 30*time.Second); err != nil {
return err.Error(), err
}
// submit 1000 activities
var futures []workflow.Future
for i:=0;i<1000;i++ {
f := workflow.ExecuteActivity(wfCtx, w.Act, req, i)
futures = append(futures, f)
}
// Wait for all tasks to run successfully
for _, f := range futures {
res := new(Result)
err := f.Get(wfCtx, res)
if err != nil {
return err.Error(), err
}
if res.KnownErr != "" {
logger.Info("known error: %v, workflow can return directly")
return res.KnownErr, nil
}
}
}
the activity options as follow:
func (w *Worker) GetActivityContext(wfCtx workflow.Context) workflow.Context {
return workflow.WithActivityOptions(wfCtx, workflow.ActivityOptions{
ScheduleToCloseTimeout: time.Hour,
StartToCloseTimeout: time.Hour,
WaitForCancellation: false,
RetryPolicy: &temporal.RetryPolicy{
MaximumAttempts: 20,
NonRetryableErrorTypes: []string{},
},
})
}
the workflow options:
"start_workflow_options": {
"workflow_run_timeout": "1d",
"workflow_task_timeout": "90m",
"retry_policy": {
"initial_interval": "1s",
"backoff_coefficient": 2.0,
"maximum_interval": "1h",
"maximum_attempts": 3
}
}
the worker options:
"options": {
"max_concurrent_activity_execution_size": 10
}
Actual Behavior
It reports WorkflowTaskTimeOut, and none of the activities was scheduled.
The timeline is as follow:
- WorkflowExecutionStarted
- WorkflowTaskScheduled
- WorkflowTaskStarted
- WorkflowTaskCompleted
- TimerStarted
- TimerFired
- WorkflowTaskScheduled (happend at time T)
- WorkflowTaskStarted
- WorkflowTaskTimedOut (happend at time T+2 minutes, timeout Type was StartToClose. but my activities' StartToCloseTimeout option is one hour)
p.s. if i reduce the number of activities to 100, all the activities run successfully.
Steps to Reproduce the Problem
Specifications
- Version:
- Platform:
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start with the Go workflow entry point shown in Worker.Start, especially the loop calling workflow.ExecuteActivity and the subsequent Future.Get calls. Reproduce the difference between 100 and 1,000 activities, then inspect the workflow-task timeout timeline alongside the 90-minute workflow-task and one-hour activity options. Done means identifying why the larger submission times out before activities are scheduled and defining the supported behavior or required change.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- go
- Domain
- backend, distributed-systems
- Issue type
- Bug
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100