actions/actions-runner-controller
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Prevent static runners from terminating due to unregister timeout

Browse Source 
The unregister timeout of 1 minute (no matter how long it is) can negatively impact availability of static runner constantly running workflow jobs, and ephemeral runner that runs a long-running job.
We deal with that by completely removing the unregistaration timeout, so that regarldess of the type of runner(static or ephemeral) it waits forever until it successfully to get unregistered before being terminated.
This commit is contained in:
Yusuke Kuoka
2022-03-13 07:22:04 +00:00
parent a1c6d1d11a
commit c4b24f8366
8 changed files with 29 additions and 38 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
controllers/runner_controller.go
View File

@@ -73,7 +73,6 @@ type RunnerReconciler struct {
RegistrationRecheckInterval time.Duration
RegistrationRecheckJitter time.Duration
UnregistrationTimeout time.Duration
UnregistrationRetryDelay time.Duration
}
@@ -212,7 +211,7 @@ func (r *RunnerReconciler) processRunnerDeletion(runner v1alpha1.Runner, ctx con
func (r *RunnerReconciler) processRunnerCreation(ctx context.Context, runner v1alpha1.Runner, log logr.Logger) (reconcile.Result, error) {
if updated, err := r.updateRegistrationToken(ctx, runner); err != nil {
return ctrl.Result{}, err
return ctrl.Result{RequeueAfter: RetryDelayOnCreateRegistrationError}, nil
} else if updated {
return ctrl.Result{Requeue: true}, nil
}
@@ -254,6 +253,10 @@ func (r *RunnerReconciler) updateRegistrationToken(ctx context.Context, runner v
rt, err := r.GitHubClient.GetRegistrationToken(ctx, runner.Spec.Enterprise, runner.Spec.Organization, runner.Spec.Repository, runner.Name)
if err != nil {
// An error can be a permanent, permission issue like the below:
// POST https://api.github.com/enterprises/YOUR_ENTERPRISE/actions/runners/registration-token: 403 Resource not accessible by integration []
// In such case retrying in seconds might not make much sense.
r.Recorder.Event(&runner, corev1.EventTypeWarning, "FailedUpdateRegistrationToken", "Updating registration token failed")
log.Error(err, "Failed to get new registration token")
return false, err