refactor: remove dead code and consolidate tests (#1335)

Removes dead code and redundant logic and consolidates tests, so there is less to read and maintain. No behavior change.

1. **act/runner**: inline single-use wrappers, compile regexes once instead of per call, embed the trampoline as a string, drop the unused legacy result of `parseActionCommand`
2. **act/container**: `sanitizeConfig` and `mergeContainerConfigs` stop returning the configs they already mutate in place, drop never-set option fields and the unused `user` parameter of `ExecWithCmdLine`
3. **act/common and cache server**: drop the unused `Trace`/`Tracef` loggers and the short SHA of `FindGitRevision`, `RegisterJob` releases through `RevokeJob`
4. **daemon, poller and reporter**: one shutdown wait for both poll modes, drop an unreachable fetch timeout branch, guard the reporter's last report time with its existing lock instead of an atomic
5. **build**: drop unused Makefile and release workflow variables, collapse the goreleaser ignore list to the same build targets
6. **tests**: fold near-duplicate tests into tables with shared setup

Throughout, hand-rolled logic gives way to `cmp.Or`, `strings.Cut`, `slices` and `min`/`max`.

Reviewed-on: https://gitea.com/gitea/runner/pulls/1335
Reviewed-by: bircni <[email protected]>
Co-authored-by: silverwind <[email protected]>
This commit is contained in:
silverwind authored and silverwind committed 2026-10-22 19:41:10 +00:00
1 parent 923a0bfb59
commit d882311ec6
107 files changed
+2408 -6500

No files matched your search

+17 -29
View File
@@ -83,8 +83,6 @@ func New(cfg *config.Config, client client.Client, runner TaskRunner) *Poller {
jobsCtx, shutdownJobs := context.WithCancel(context.Background())
done := make(chan struct{})
p := &Poller{
client: client,
runner: runner,
@@ -96,11 +94,11 @@ func New(cfg *config.Config, client client.Client, runner TaskRunner) *Poller {
jobsCtx: jobsCtx,
shutdownJobs: shutdownJobs,
done: done,
done: make(chan struct{}),
availabilityReady: true,
availabilityReason: "ok",
}
p.lastHealthyPoll.Store(time.Now().UnixNano())
p.availabilityReady = true
p.availabilityReason = "ok"
return p
}
@@ -132,8 +130,8 @@ func (p *Poller) Poll() {
continue
}
task, ok := p.fetchTask(p.pollingCtx, s)
if !ok {
task := p.fetchTask(p.pollingCtx, s)
if task == nil {
p.runIdleMaintenance()
<-sem
if !p.waitBackoff(s) {
@@ -166,8 +164,8 @@ func (p *Poller) PollOnce(onTaskFetched func()) {
}
continue
}
task, ok := p.fetchTask(p.pollingCtx, s)
if !ok {
task := p.fetchTask(p.pollingCtx, s)
if task == nil {
p.runIdleMaintenance()
if !p.waitBackoff(s) {
return
@@ -348,7 +346,7 @@ func (p *Poller) runTaskWithRecover(ctx context.Context, task *runnerv1.Task) {
}
}
func (p *Poller) fetchTask(ctx context.Context, s *workerState) (*runnerv1.Task, bool) {
func (p *Poller) fetchTask(ctx context.Context, s *workerState) *runnerv1.Task {
reqCtx, cancel := context.WithTimeout(ctx, p.cfg.Runner.FetchTimeout)
defer cancel()
@@ -370,7 +368,7 @@ func (p *Poller) fetchTask(ctx context.Context, s *workerState) (*runnerv1.Task,
}
s.consecutiveEmpty++
metrics.PollFetchTotal.WithLabelValues(metrics.LabelResultEmpty).Inc()
return nil, false
return nil
}
s.fetchTimedOut = false
metrics.PollFetchDuration.Observe(time.Since(start).Seconds())
@@ -385,22 +383,17 @@ func (p *Poller) fetchTask(ctx context.Context, s *workerState) (*runnerv1.Task,
p.unregistered.Store(true)
p.registrationRejected.Store(rejected)
p.shutdownPolling()
return nil, false
}
// Not a long poll, so a deadline can mean the server assigned a task
// this runner never received.
if errors.Is(err, context.DeadlineExceeded) {
logging.WithError(err).Errorf("fetching a task timed out after %s", p.cfg.Runner.FetchTimeout)
} else {
logging.WithError(err).Error("failed to fetch task")
return nil
}
logging.WithError(err).Error("failed to fetch task")
p.lastPollFailed.Store(true)
s.consecutiveErrors++
metrics.PollFetchTotal.WithLabelValues(metrics.LabelResultError).Inc()
metrics.ClientErrors.WithLabelValues(metrics.LabelMethodFetchTask, connect.CodeOf(err).String()).Inc()
return nil, false
return nil
}
p.markHealthyPoll()
p.lastHealthyPoll.Store(time.Now().UnixNano())
p.lastPollFailed.Store(false)
// Successful response — reset error counter.
s.consecutiveErrors = 0
@@ -408,7 +401,7 @@ func (p *Poller) fetchTask(ctx context.Context, s *workerState) (*runnerv1.Task,
if resp == nil {
s.consecutiveEmpty++
metrics.PollFetchTotal.WithLabelValues(metrics.LabelResultEmpty).Inc()
return nil, false
return nil
}
if resp.TasksVersion > v {
@@ -418,17 +411,12 @@ func (p *Poller) fetchTask(ctx context.Context, s *workerState) (*runnerv1.Task,
if resp.Task == nil {
s.consecutiveEmpty++
metrics.PollFetchTotal.WithLabelValues(metrics.LabelResultEmpty).Inc()
return nil, false
return nil
}
// got a task, set `tasksVersion` to zero to force query db in next request.
p.tasksVersion.CompareAndSwap(resp.TasksVersion, 0)
metrics.PollFetchTotal.WithLabelValues(metrics.LabelResultTask).Inc()
return resp.Task, true
}
func (p *Poller) markHealthyPoll() {
p.lastHealthyPoll.Store(time.Now().UnixNano())
p.lastPollFailed.Store(false)
return resp.Task
}
+42 -100
View File
@@ -22,14 +22,10 @@ import (
"github.com/stretchr/testify/require"
)
// TestPoller_WorkerStateCounters verifies that workerState correctly tracks
// consecutive empty responses independently per state instance, and that
// fetchTask increments only the relevant counter.
func TestPoller_WorkerStateCounters(t *testing.T) {
client := mocks.NewClient(t)
client.On("FetchTask", mock.Anything, mock.Anything).Return(
func(_ context.Context, _ *runnerv1.FetchTaskRequest) (*runnerv1.FetchTaskResponse, error) {
// Always return an empty response.
return &runnerv1.FetchTaskResponse{}, nil
},
)
@@ -42,24 +38,17 @@ func TestPoller_WorkerStateCounters(t *testing.T) {
s1 := &workerState{}
s2 := &workerState{}
// Each worker independently observes one empty response.
_, ok := p.fetchTask(ctx, s1)
require.False(t, ok)
_, ok = p.fetchTask(ctx, s2)
require.False(t, ok)
require.Nil(t, p.fetchTask(ctx, s1))
require.Nil(t, p.fetchTask(ctx, s2))
assert.Equal(t, int64(1), s1.consecutiveEmpty, "worker 1 should only count its own empty response")
assert.Equal(t, int64(1), s2.consecutiveEmpty, "worker 2 should only count its own empty response")
// Worker 1 sees a second empty; worker 2 stays at 1.
_, ok = p.fetchTask(ctx, s1)
require.False(t, ok)
require.Nil(t, p.fetchTask(ctx, s1))
assert.Equal(t, int64(2), s1.consecutiveEmpty)
assert.Equal(t, int64(1), s2.consecutiveEmpty, "worker 2's counter must not be affected by worker 1's empty fetches")
}
// TestPoller_FetchErrorIncrementsErrorsOnly verifies that a fetch error
// increments only the per-worker error counter, not the empty counter.
func TestPoller_FetchErrorIncrementsErrorsOnly(t *testing.T) {
client := mocks.NewClient(t)
client.On("FetchTask", mock.Anything, mock.Anything).Return(nil, errors.New("network unreachable"))
@@ -69,15 +58,11 @@ func TestPoller_FetchErrorIncrementsErrorsOnly(t *testing.T) {
p := &Poller{client: client, cfg: cfg}
s := &workerState{}
_, ok := p.fetchTask(t.Context(), s)
require.False(t, ok)
require.Nil(t, p.fetchTask(t.Context(), s))
assert.Equal(t, int64(1), s.consecutiveErrors)
assert.Equal(t, int64(0), s.consecutiveEmpty)
}
// A deadline is the idle path once the server holds the request open, and a
// symptom before then, so it must neither reset the error count nor assert that
// the server is reachable.
func TestPoller_FetchTimeoutIsNoSignal(t *testing.T) {
client := mocks.NewClient(t)
client.On("FetchTask", mock.Anything, mock.Anything).Return(nil, context.DeadlineExceeded)
@@ -90,20 +75,15 @@ func TestPoller_FetchTimeoutIsNoSignal(t *testing.T) {
hook := logtest.SetDefault(t)
s := &workerState{consecutiveErrors: 2}
_, ok := p.fetchTask(t.Context(), s)
require.False(t, ok)
require.Nil(t, p.fetchTask(t.Context(), s))
assert.Equal(t, int64(2), s.consecutiveErrors)
assert.Equal(t, int64(1), s.consecutiveEmpty)
assert.True(t, p.lastPollFailed.Load(), "a timeout must not clear a known failure")
// An idle runner against a server holding the request open must not flood.
_, _ = p.fetchTask(t.Context(), s)
p.fetchTask(t.Context(), s)
assert.Len(t, hook.AllRecords(), 1)
}
// TestPoller_FetchUnauthenticatedStopsPolling verifies that an Unauthenticated
// response marks the runner as unregistered and cancels the polling context so
// the daemon can exit instead of retrying forever.
func TestPoller_FetchUnauthenticatedStopsPolling(t *testing.T) {
for name, tc := range map[string]struct {
fetchErr error
@@ -120,8 +100,7 @@ func TestPoller_FetchUnauthenticatedStopsPolling(t *testing.T) {
require.NoError(t, err)
p := New(cfg, client, nil)
s := &workerState{}
_, ok := p.fetchTask(t.Context(), s)
require.False(t, ok)
require.Nil(t, p.fetchTask(t.Context(), s))
assert.True(t, p.Unregistered(), "runner should be marked unregistered")
assert.Equal(t, tc.rejected, p.RegistrationRejected())
@@ -135,8 +114,6 @@ func TestPoller_FetchUnauthenticatedStopsPolling(t *testing.T) {
}
}
// TestPoller_CalculateInterval verifies the exponential backoff math is
// correctly driven by the workerState counters.
func TestPoller_CalculateInterval(t *testing.T) {
cfg, err := config.LoadDefault("")
require.NoError(t, err)
@@ -165,7 +142,6 @@ func TestPoller_CalculateInterval(t *testing.T) {
}
}
// atomicMax atomically updates target to max(target, val).
func atomicMax(target *atomic.Int64, val int64) {
for {
old := target.Load()
@@ -205,7 +181,6 @@ func TestPollerReady(t *testing.T) {
require.NoError(t, err)
poller := New(cfg, nil, &availabilityRunner{})
// /readyz reuses the availability the poll loop last recorded.
ready, reason := poller.Ready(time.Second)
assert.True(t, ready)
assert.Equal(t, "ok", reason)
@@ -291,63 +266,45 @@ func (m *idleAwareRunner) OnIdle(_ context.Context) {
m.idleCalls.Add(1)
}
func TestPollerPollCallsOnIdle(t *testing.T) {
cli := mocks.NewClient(t)
cli.On("FetchTask", mock.Anything, mock.Anything).Return(
func(_ context.Context, _ *runnerv1.FetchTaskRequest) (*runnerv1.FetchTaskResponse, error) {
return &runnerv1.FetchTaskResponse{}, nil
},
)
func TestPollerCallsOnIdle(t *testing.T) {
for _, pollOnce := range []bool{false, true} {
name := "Poll"
if pollOnce {
name = "PollOnce"
}
t.Run(name, func(t *testing.T) {
cli := mocks.NewClient(t)
cli.On("FetchTask", mock.Anything, mock.Anything).Return(&runnerv1.FetchTaskResponse{}, nil)
cfg, err := config.LoadDefault("")
require.NoError(t, err)
cfg.Runner.Capacity = 1
cfg.Runner.FetchInterval = 10 * time.Millisecond
cfg.Runner.FetchIntervalMax = 10 * time.Millisecond
cfg, err := config.LoadDefault("")
require.NoError(t, err)
if !pollOnce {
cfg.Runner.Capacity = 1
}
cfg.Runner.FetchInterval = 10 * time.Millisecond
cfg.Runner.FetchIntervalMax = 10 * time.Millisecond
runner := &idleAwareRunner{}
poller := New(cfg, cli, runner)
runner := &idleAwareRunner{}
poller := New(cfg, cli, runner)
var wg sync.WaitGroup
wg.Go(func() {
if pollOnce {
poller.PollOnce(func() {})
} else {
poller.Poll()
}
})
var wg sync.WaitGroup
wg.Go(poller.Poll)
require.Eventually(t, func() bool {
return runner.idleCalls.Load() > 0
}, time.Second, 10*time.Millisecond)
require.Eventually(t, func() bool {
return runner.idleCalls.Load() > 0
}, time.Second, 10*time.Millisecond)
ctx, cancel := context.WithTimeout(t.Context(), time.Second)
defer cancel()
require.NoError(t, poller.Shutdown(ctx))
wg.Wait()
}
func TestPollerPollOnceCallsOnIdle(t *testing.T) {
cli := mocks.NewClient(t)
cli.On("FetchTask", mock.Anything, mock.Anything).Return(
func(_ context.Context, _ *runnerv1.FetchTaskRequest) (*runnerv1.FetchTaskResponse, error) {
return &runnerv1.FetchTaskResponse{}, nil
},
)
cfg, err := config.LoadDefault("")
require.NoError(t, err)
cfg.Runner.FetchInterval = 10 * time.Millisecond
cfg.Runner.FetchIntervalMax = 10 * time.Millisecond
runner := &idleAwareRunner{}
poller := New(cfg, cli, runner)
var wg sync.WaitGroup
wg.Go(func() { poller.PollOnce(func() {}) })
require.Eventually(t, func() bool {
return runner.idleCalls.Load() > 0
}, time.Second, 10*time.Millisecond)
ctx, cancel := context.WithTimeout(t.Context(), time.Second)
defer cancel()
require.NoError(t, poller.Shutdown(ctx))
wg.Wait()
ctx, cancel := context.WithTimeout(t.Context(), time.Second)
defer cancel()
require.NoError(t, poller.Shutdown(ctx))
wg.Wait()
})
}
}
func TestPollerPollOnceCallsOnTaskFetchedBeforeRunningIt(t *testing.T) {
@@ -362,9 +319,6 @@ func TestPollerPollOnceCallsOnTaskFetchedBeforeRunningIt(t *testing.T) {
assert.Equal(t, int64(1), runner.totalCompleted.Load())
}
// TestPoller_ConcurrencyLimitedByCapacity verifies that with capacity=3 and
// 6 available tasks, at most 3 tasks run concurrently, and FetchTask is
// never called concurrently (single poller).
func TestPoller_ConcurrencyLimitedByCapacity(t *testing.T) {
const (
capacity = 3
@@ -427,13 +381,6 @@ func TestPoller_ConcurrencyLimitedByCapacity(t *testing.T) {
"all tasks should have been executed")
}
// TestPoller_ShutdownForcesJobsOnTimeout locks in the fix for a
// pre-existing bug where Shutdown's timeout branch used a blocking
// `<-p.done` receive, leaving p.shutdownJobs() unreachable. With a
// task parked on jobsCtx and a Shutdown deadline shorter than the
// task's natural completion, Shutdown must force-cancel via
// shutdownJobs() and return ctx.Err() promptly — not block until the
// task would have finished on its own.
func TestPoller_ShutdownForcesJobsOnTimeout(t *testing.T) {
var served atomic.Bool
cli := mocks.NewClient(t)
@@ -448,8 +395,6 @@ func TestPoller_ShutdownForcesJobsOnTimeout(t *testing.T) {
},
)
// delay >> Shutdown timeout: Run only returns when jobsCtx is
// cancelled by shutdownJobs().
runner := &mockRunner{delay: 30 * time.Second}
cfg, err := config.LoadDefault("")
@@ -474,9 +419,6 @@ func TestPoller_ShutdownForcesJobsOnTimeout(t *testing.T) {
elapsed := time.Since(start)
require.ErrorIs(t, err, context.DeadlineExceeded)
// With the fix, Shutdown returns shortly after the deadline once
// the forced job unwinds. Without the fix, the blocking <-p.done
// would hang for the full 30s mockRunner delay.
assert.Less(t, elapsed, 5*time.Second,
"Shutdown must not block on the parked task; shutdownJobs() must run on timeout")