s3 lifecycle: count filer entries a walk reads but skips as progress (#12658)

* dailyrun: count entries read from the filer, skipped ones included

A versioned bucket's pre-pass and a resumed walk read entries without
emitting them, so a walk can read for hours with no visited entry.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

* s3 lifecycle: count listed entries as progress work

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>

weed/s3api/s3lifecycle/dailyrun/run_stats.go

@@ -1,10 +1,14 @@
 package dailyrun
 
 import (
+	"context"
 	"strings"
 	"sync"
 	"sync/atomic"
 
+	"google.golang.org/grpc"
+
+	"github.com/seaweedfs/seaweedfs/weed/pb/filer_pb"
 	"github.com/seaweedfs/seaweedfs/weed/stats"
 )
 
@@ -46,7 +50,10 @@ type RunSummary struct {
 	EntriesVisited int64
 	// Distinct object keys among the listed entries.
 	ObjectsVisited int64
-	EventsScanned  int64
+	// Every entry read from the filer, including the ones a walk skips
+	// without visiting (versioned pre-pass, resume start, other shards).
+	EntriesListed int64
+	EventsScanned int64
 	// LifecycleDelete outcomes keyed by the dispatch_total outcome label.
 	// They count RPC results, not reclaimed space.
 	Outcomes    map[string]int64
@@ -63,6 +70,7 @@ type RunSummary struct {
 type RunStats struct {
 	entries atomic.Int64
 	objects atomic.Int64
+	listed  atomic.Int64
 	events  atomic.Int64
 
 	mu           sync.Mutex
@@ -123,6 +131,41 @@ func (s *RunStats) visitEntry(newObject bool) {
 	}
 }
 
+// CountListed returns client with its ListEntries results counted as
+// EntriesListed.
+func (s *RunStats) CountListed(client filer_pb.SeaweedFilerClient) filer_pb.SeaweedFilerClient {
+	if s == nil || client == nil {
+		return client
+	}
+	return &listCountingClient{SeaweedFilerClient: client, stats: s}
+}
+
+type listCountingClient struct {
+	filer_pb.SeaweedFilerClient
+	stats *RunStats
+}
+
+func (c *listCountingClient) ListEntries(ctx context.Context, in *filer_pb.ListEntriesRequest, opts ...grpc.CallOption) (filer_pb.SeaweedFiler_ListEntriesClient, error) {
+	stream, err := c.SeaweedFilerClient.ListEntries(ctx, in, opts...)
+	if err != nil {
+		return nil, err
+	}
+	return &listCountingStream{SeaweedFiler_ListEntriesClient: stream, stats: c.stats}, nil
+}
+
+type listCountingStream struct {
+	filer_pb.SeaweedFiler_ListEntriesClient
+	stats *RunStats
+}
+
+func (s *listCountingStream) Recv() (*filer_pb.ListEntriesResponse, error) {
+	resp, err := s.SeaweedFiler_ListEntriesClient.Recv()
+	if err == nil && resp.GetEntry() != nil {
+		s.stats.listed.Add(1)
+	}
+	return resp, err
+}
+
 func (s *RunStats) scanEvent() {
 	if s == nil {
 		return
@@ -176,6 +219,7 @@ func (s *RunStats) Summary() RunSummary {
 	out := RunSummary{
 		EntriesVisited: s.entries.Load(),
 		ObjectsVisited: s.objects.Load(),
+		EntriesListed:  s.listed.Load(),
 		EventsScanned:  s.events.Load(),
 	}
 	s.mu.Lock()

weed/s3api/s3lifecycle/dailyrun/run_stats_test.go

@@ -7,6 +7,7 @@ import (
 	"testing"
 	"time"
 
+	"github.com/seaweedfs/seaweedfs/weed/pb/filer_pb"
 	"github.com/seaweedfs/seaweedfs/weed/pb/s3_lifecycle_pb"
 	"github.com/seaweedfs/seaweedfs/weed/s3api/s3lifecycle"
 	"github.com/seaweedfs/seaweedfs/weed/s3api/s3lifecycle/bootstrap"
@@ -167,3 +168,24 @@ func TestRunStats_NilIsANoOp(t *testing.T) {
 	runStats.setShardErrors(1)
 	assert.Equal(t, RunSummary{}, runStats.Summary())
 }
+
+// A walk can read a bucket without visiting anything: the versioned
+// pre-pass reads plain files without emitting them, and a resumed walk
+// skips every key up to its start.
+func TestRunStats_CountListedCountsSkippedEntries(t *testing.T) {
+	mtime := time.Now().Add(-7 * 24 * time.Hour)
+	client := &fakeFiler{tree: map[string][]*filer_pb.Entry{
+		"/buckets/bkt": {file("a.txt", mtime, 1), file("b.txt", mtime, 1), file("c.txt", mtime, 1)},
+	}}
+	stats := &RunStats{}
+	emitted := 0
+	require.NoError(t, FilerListFunc(stats.CountListed(client), "/buckets", nil)(context.Background(), "bkt", "c.txt", func(*bootstrap.Entry) error {
+		emitted++
+		return nil
+	}))
+
+	assert.Zero(t, emitted)
+	assert.Equal(t, int64(6), stats.Summary().EntriesListed, "both passes over the versioned bucket")
+	var nilStats *RunStats
+	assert.Same(t, client, nilStats.CountListed(client))
+}

weed/worker/tasks/s3_lifecycle/handler.go

@@ -337,7 +337,7 @@ func (h *Handler) executeDailyReplay(ctx context.Context, request *plugin_pb.Exe
 	walkerDispatch := &dailyrun.WalkerDispatcher{Client: client, Limiter: limiter, Stats: runStats}
 	walker := dailyrun.WalkerFunc(func(walkCtx context.Context, view *engine.Snapshot, shards []int) error {
 		return dailyrun.WalkBuckets(walkCtx, view, shards, buckets,
-			dailyrun.FilerListFunc(filerClient, bucketsPath, view), walkerDispatch, runStats)
+			dailyrun.FilerListFunc(runStats.CountListed(filerClient), bucketsPath, view), walkerDispatch, runStats)
 	})
 
 	_ = sender.SendProgress(&plugin_pb.JobProgressUpdate{

weed/worker/tasks/s3_lifecycle/progress.go

@@ -15,7 +15,7 @@ import (
 const progressInterval = 5 * time.Minute
 
 func runWork(s dailyrun.RunSummary) int64 {
-	work := s.EntriesVisited + s.EventsScanned
+	work := s.EntriesVisited + s.EntriesListed + s.EventsScanned
 	for _, n := range s.Outcomes {
 		work += n
 	}
@@ -23,8 +23,8 @@ func runWork(s dailyrun.RunSummary) int64 {
 }
 
 func progressMessage(s dailyrun.RunSummary) string {
-	return fmt.Sprintf("s3 lifecycle daily replay running: last_phase=%s last_bucket=%s entries=%d objects=%d events=%d rpc_done=%d rpc_errors=%d",
-		s.Phase, s.Bucket, s.EntriesVisited, s.ObjectsVisited, s.EventsScanned, s.Outcomes["DONE"], s.Outcomes["RPC_ERROR"])
+	return fmt.Sprintf("s3 lifecycle daily replay running: last_phase=%s last_bucket=%s entries=%d objects=%d listed=%d events=%d rpc_done=%d rpc_errors=%d",
+		s.Phase, s.Bucket, s.EntriesVisited, s.ObjectsVisited, s.EntriesListed, s.EventsScanned, s.Outcomes["DONE"], s.Outcomes["RPC_ERROR"])
 }
 
 // startProgressReporter returns stop, which returns only once no further

weed/worker/tasks/s3_lifecycle/progress_test.go

@@ -46,7 +46,7 @@ func (s *progressSender) tries() int {
 }
 
 type fakeRun struct {
-	entries, events, done, rpcErrors atomic.Int64
+	entries, listed, events, done, rpcErrors atomic.Int64
 }
 
 func (f *fakeRun) summary() dailyrun.RunSummary {
@@ -54,6 +54,7 @@ func (f *fakeRun) summary() dailyrun.RunSummary {
 		Phase:          dailyrun.PhaseWalk,
 		Bucket:         "b1",
 		EntriesVisited: f.entries.Load(),
+		EntriesListed:  f.listed.Load(),
 		EventsScanned:  f.events.Load(),
 		Outcomes:       map[string]int64{"DONE": f.done.Load(), "RPC_ERROR": f.rpcErrors.Load()},
 	}
@@ -80,7 +81,7 @@ func TestProgressReporter_ReportsEachKindOfWork(t *testing.T) {
 	stop := startProgressReporter(progressRequest(), sender, run.summary, testInterval)
 	defer stop()
 
-	bumps := []*atomic.Int64{&run.entries, &run.events, &run.done, &run.rpcErrors}
+	bumps := []*atomic.Int64{&run.entries, &run.listed, &run.events, &run.done, &run.rpcErrors}
 	for i, counter := range bumps {
 		counter.Add(1)
 		require.Eventually(t, func() bool { return len(sender.sent()) == i+1 }, time.Second, testInterval)
@@ -96,6 +97,7 @@ func TestProgressReporter_ReportsEachKindOfWork(t *testing.T) {
 	require.Equal(t, "running", last.Stage)
 	require.Contains(t, last.Message, "last_phase=walk last_bucket=b1")
 	require.Contains(t, last.Message, "entries=1")
+	require.Contains(t, last.Message, "listed=1")
 	require.Contains(t, last.Message, "events=1")
 	require.Contains(t, last.Message, "rpc_done=1")
 	require.Contains(t, last.Message, "rpc_errors=1")