enhance(config): add container.sweep option (#1275)

The Docker proxy labels every container, network and volume a job creates and removes them when the job ends, which breaks jobs that deploy long-running services to the runner's own Docker daemon. Document that behavior and add `container.sweep` (default `true`) so operators who deploy to the same daemon can opt out. Kept resources stay labeled with `com.gitea.runner.job`.

Fixes: https://gitea.com/gitea/runner/issues/1243
Co-authored-by: silverwind <[email protected]>
Reviewed-on: https://gitea.com/gitea/runner/pulls/1275
Reviewed-by: silverwind <[email protected]>
Co-authored-by: bircni <[email protected]>

act/runner/run_context.go

@@ -725,7 +725,7 @@ func (rc *RunContext) cleanupJobResources(networkName string, createAndDeleteNet
 			logger.Infof("Cleaning up services for job %s", rc.JobName)
 			errs = append(errs, rc.stopServiceContainers()(ctx))
 		}
-		if !common.Dryrun(ctx) && (preclean || rc.hadDockerProxy) {
+		if !common.Dryrun(ctx) && !rc.Config.NoSweep && (preclean || rc.hadDockerProxy) {
 			errs = append(errs, container.RemoveDockerJobResources(ctx, rc.jobContainerName()))
 		}
 		if preclean || rc.deferVolumeCleanup == nil {

act/runner/run_context_test.go

@@ -778,9 +778,10 @@ func TestRemoveDockerActionImagesByTag(t *testing.T) {
 func TestCleanupJobResourcesContinuesAfterFailure(t *testing.T) {
 	t.Setenv("TMPDIR", "/tmp")
 	proxyDir := t.TempDir()
-	for _, name := range []string{"synchronous", "proxy", "closed proxy", "deferred", "preclean deferred", "canceled"} {
+	for _, name := range []string{"synchronous", "proxy", "closed proxy", "kept proxy", "deferred", "preclean deferred", "canceled"} {
 		t.Run(name, func(t *testing.T) {
 			proxy, preclean, deferred := strings.HasSuffix(name, "proxy"), strings.HasPrefix(name, "preclean"), strings.HasSuffix(name, "deferred")
+			removesLabelled := (proxy || preclean) && name != "kept proxy"
 			if proxy && runtime.GOOS == "windows" {
 				t.Skip("Unix socket ownership is unavailable on Windows")
 			}
@@ -790,7 +791,7 @@ func TestCleanupJobResourcesContinuesAfterFailure(t *testing.T) {
 			service.On("Remove").Return(func(context.Context) error { return removeError }).Once()
 			service.On("Close").Return(func(context.Context) error { return closeError }).Once()
 			rc := &RunContext{
-				Config:            &Config{CacheContainer: "cache-container"},
+				Config:            &Config{CacheContainer: "cache-container", NoSweep: name == "kept proxy"},
 				Run:               &model.Run{Workflow: &model.Workflow{Name: "wf"}, JobID: "job"},
 				JobContainer:      job,
 				serviceContainers: []*serviceContainer{{name: "svc", container: service}},
@@ -846,8 +847,8 @@ func TestCleanupJobResourcesContinuesAfterFailure(t *testing.T) {
 				}
 				assert.Equal(t, 2, volumeRemovals)
 			}
-			assert.Equal(t, proxy || preclean, strings.Contains(err.Error(), "labelled GET /containers/json"))
-			if proxy || preclean {
+			assert.Equal(t, removesLabelled, strings.Contains(err.Error(), "labelled GET /containers/json"))
+			if removesLabelled {
 				require.ErrorContains(t, err, "labelled GET /networks")
 				require.ErrorContains(t, err, "labelled GET /volumes")
 			}

act/runner/runner.go

@@ -87,6 +87,7 @@ type Config struct {
 	MaxParallel                       int           // max parallel jobs to run across all workflows (0 = no limit, uses CPU count)
 	AllocatePTY                       bool          // allocate a pseudo-TTY for each step's process
 	ServiceReadyTimeout               time.Duration // how long a job waits for its service containers to report healthy (0 uses the default)
+	NoSweep                           bool          // keep the containers, networks and volumes a job creates through its docker proxy
 	RunnerName                        string        // name this runner registered with, reported as `runner.name`, defaults to the hostname
 	JobStartedHook                    string        // script run inside the job environment before the job's first step; ACTIONS_RUNNER_HOOK_JOB_STARTED is read from Env when empty
 	JobCompletedHook                  string        // script run inside the job environment after the job's last step; ACTIONS_RUNNER_HOOK_JOB_COMPLETED is read from Env when empty