Merge pull request #8937 from dokku/8912-cron-containers-are-never-retired-past-their-active-deadline

Retire cron containers past their active deadline

plugins/cron/cron.go

@@ -42,6 +42,23 @@ var (
 
 const MaintenancePropertyPrefix = "maintenance."
 
+// DefaultTTLSeconds is how long a cron task may run before it is reaped. The
+// docker-local scheduler stamps this onto the container as the
+// com.dokku.active-deadline-seconds label, while the k3s scheduler renders it
+// as the CronJob's activeDeadlineSeconds.
+const DefaultTTLSeconds int64 = 86400
+
+// validateTTLSeconds returns an error if the requested task lifetime is not a
+// positive number of seconds. A zero or negative deadline would either expire
+// the task the instant it starts or never expire it at all.
+func validateTTLSeconds(ttlSeconds int64) error {
+	if ttlSeconds <= 0 {
+		return fmt.Errorf("--ttl-seconds must be a positive integer")
+	}
+
+	return nil
+}
+
 // CronTask is a struct that represents a cron task
 type CronTask struct {
 	// ID is a unique identifier for the cron task

plugins/cron/cron_test.go

@@ -123,3 +123,32 @@ func TestDokkuRunCommandAppTaskIgnoresLogFile(t *testing.T) {
 		t.Errorf("DokkuRunCommand() interpolated a redirect into an app task line: %q", got)
 	}
 }
+
+// TestValidateTTLSeconds pins that a cron task deadline must be a positive
+// number of seconds. A zero or negative value would either expire the task the
+// instant it starts or leave it running forever, and the docker-local retire
+// pass only reaps containers whose deadline has actually elapsed.
+func TestValidateTTLSeconds(t *testing.T) {
+	cases := []struct {
+		name       string
+		ttlSeconds int64
+		wantErr    bool
+	}{
+		{name: "default", ttlSeconds: DefaultTTLSeconds},
+		{name: "positive override", ttlSeconds: 1},
+		{name: "zero", ttlSeconds: 0, wantErr: true},
+		{name: "negative", ttlSeconds: -1, wantErr: true},
+	}
+
+	for _, tc := range cases {
+		t.Run(tc.name, func(t *testing.T) {
+			err := validateTTLSeconds(tc.ttlSeconds)
+			if tc.wantErr && err == nil {
+				t.Errorf("validateTTLSeconds(%d) = nil, want an error", tc.ttlSeconds)
+			}
+			if !tc.wantErr && err != nil {
+				t.Errorf("validateTTLSeconds(%d) = %v, want nil", tc.ttlSeconds, err)
+			}
+		})
+	}
+}

plugins/cron/src/commands/commands.go

@@ -21,7 +21,7 @@ Additional commands:`
     cron:list <app> [--format json|stdout], List scheduled cron tasks for an app
     cron:report [<app>] [<flag>], Display report about an app
     cron:resume <app> <cron_id>, Resume a cron task
-    cron:run <app> <cron_id> [--detach], Run a cron task on the fly
+    cron:run <app> <cron_id> [--detach] [--ttl-seconds SECONDS], Run a cron task on the fly
     cron:set [--global|<app>] <key> <value>, Set or clear a cron property for an app
     cron:suspend <app> <cron_id>, Suspend a cron task`
 )

plugins/cron/src/subcommands/subcommands.go

@@ -49,10 +49,11 @@ func main() {
 	case "run":
 		args := flag.NewFlagSet("cron:run", flag.ExitOnError)
 		detached := args.Bool("detach", false, "--detach: run the container in a detached mode")
+		ttlSeconds := args.Int64("ttl-seconds", cron.DefaultTTLSeconds, "--ttl-seconds: number of seconds the task may run before it is reaped")
 		args.Parse(os.Args[2:])
 		appName := args.Arg(0)
 		cronID := args.Arg(1)
-		err = cron.CommandRun(appName, cronID, *detached)
+		err = cron.CommandRun(appName, cronID, *detached, *ttlSeconds)
 	case "set":
 		args := flag.NewFlagSet("cron:set", flag.ExitOnError)
 		global := args.Bool("global", false, "--global: set a global property")

plugins/cron/subcommands.go

@@ -5,6 +5,7 @@ import (
 	"errors"
 	"fmt"
 	"os"
+	"strconv"
 	"strings"
 
 	"github.com/dokku/dokku/plugins/common"
@@ -97,11 +98,15 @@ func CommandResume(appName string, cronID string) error {
 }
 
 // CommandRun executes a cron task on the fly
-func CommandRun(appName string, cronID string, detached bool) error {
+func CommandRun(appName string, cronID string, detached bool, ttlSeconds int64) error {
 	if err := common.VerifyAppName(appName); err != nil {
 		return err
 	}
 
+	if err := validateTTLSeconds(ttlSeconds); err != nil {
+		return err
+	}
+
 	tasks, err := FetchCronTasks(FetchCronTasksInput{AppName: appName})
 	if err != nil {
 		return err
@@ -139,7 +144,7 @@ func CommandRun(appName string, cronID string, detached bool) error {
 	os.Setenv("DOKKU_CONCURRENCY_POLICY", concurrencyPolicy)
 	os.Setenv("DOKKU_CRON_ID", cronID)
 	os.Setenv("DOKKU_RM_CONTAINER", "1")
-	os.Setenv("DOKKU_RUN_TTL_SECONDS", "86400")
+	os.Setenv("DOKKU_RUN_TTL_SECONDS", strconv.FormatInt(ttlSeconds, 10))
 	scheduler := common.GetAppScheduler(appName)
 	args := append([]string{scheduler, appName, "0", "--"}, fields...)
 	_, err = common.CallPlugnTrigger(common.PlugnTriggerInput{