fix(ec): bring ec.encode worker and EC/volume helpers to parity with shell (#9599)

* refactor(volume): extract replica sync/select into shared volume_replica package

Move the volume replica reconciliation helpers (status, union builder,
SyncAndSelectBestReplica, ReadNeedleMeta) out of the shell into a new
weed/storage/volume_replica package so both the shell (ec.encode, volume.tier.move,
volume.check.disk) and the EC encode worker can reuse them. No behavior change.

* fix(ec): bring ec.encode worker to parity with the shell

- Sync replicas and encode the most-complete one (via the shared
  volume_replica.SyncAndSelectBestReplica) instead of a possibly-stale replica,
  marking all replicas readonly first. Prevents silent data loss when a stale
  replica is encoded and the originals deleted.
- Skip remote/tiered volumes in detection (shell ec.encode excludes them).
- Min-node safety gate: refuse to encode when cluster nodes < parity shards.
- Align default thresholds with the shell (fullness 0.95, quiet 1h).

* fix(vacuum): plugin path honors min_volume_age_seconds override

deriveVacuumConfig hard-coded MinVolumeAgeSeconds=0, dropping any configured
value. Read it from worker config (default 0, matching the shell/master vacuum
which has no age gate) so an explicit override is honored.

* address review feedback

- config.go: align GetConfigSpec schema defaults (quiet_for_seconds=3600,
  fullness_ratio=0.95) with the runtime defaults so UI/bootstrap flows match the
  shell (coderabbitai).
- ec_task.go: roll back readonly when markReplicasReadonly fails partway, so
  already-marked replicas don't stay readonly (coderabbitai).
- volume_replica: pass the caller's replica statuses into buildUnionReplica instead
  of re-fetching them, and skip the per-needle ReadNeedleMeta RPC when the source
  replica is read-only (gemini-code-assist).

* test(plugin_workers/ec): make fixtures eligible under the new defaults

The default EC encode thresholds were raised to match the shell (fullness 0.95,
quiet 1h), but the plugin-worker integration fixtures still used 90%-full /
10-minute-old volumes, so detection found no eligible volumes and the tests failed
in CI. Bump the eligible fixtures to 96% full and 2h old.

weed/worker/tasks/erasure_coding/detection.go

@@ -52,12 +52,19 @@ func Detection(ctx context.Context, metrics []*types.VolumeHealthMetrics, cluste
 	skippedCollectionFilter := 0
 	skippedQuietTime := 0
 	skippedFullness := 0
+	skippedRemote := 0
+	skippedTooFewNodes := 0
 	consecutivePlanningFailures := 0
 
 	var planner *ecPlacementPlanner
 
 	allowedCollections := wildcard.CompileWildcardMatchers(ecConfig.CollectionFilter)
 
+	// Cluster node count for the min-node safety gate (mirrors the shell ec.encode
+	// guard that refuses to encode when nodes < parity shards, so shards cannot be
+	// spread for fault tolerance).
+	clusterNodeCount := countTopologyNodes(clusterInfo.ActiveTopology)
+
 	// Group metrics by VolumeID to handle replicas and select canonical server
 	volumeGroups := make(map[uint32][]*types.VolumeHealthMetrics)
 	for _, metric := range metrics {
@@ -163,6 +170,21 @@ func Detection(ctx context.Context, metrics []*types.VolumeHealthMetrics, cluste
 			continue
 		}
 
+		// Skip remote/tiered volumes: encoding them would lose the tiering. The
+		// shell ec.encode excludes remote volumes for the same reason.
+		if metric.HasRemoteCopy {
+			skippedRemote++
+			continue
+		}
+
+		// Min-node safety gate: don't encode when the cluster has fewer nodes than
+		// this collection's parity shards — shards could not be spread to tolerate
+		// failures. Mirrors the shell ec.encode guard (node count < parity shards).
+		if clusterNodeCount > 0 && clusterNodeCount < erasure_coding.ParityShardsCount {
+			skippedTooFewNodes++
+			continue
+		}
+
 		// Check quiet duration and fullness criteria
 		if metric.Age >= quietThreshold && metric.FullnessRatio >= ecConfig.FullnessRatio {
 			if ctx != nil {
@@ -373,8 +395,8 @@ func Detection(ctx context.Context, metrics []*types.VolumeHealthMetrics, cluste
 	// Log debug summary if no tasks were created
 	if len(results) == 0 && len(metrics) > 0 && !stoppedEarly {
 		totalVolumes := len(metrics)
-		glog.V(1).Infof("EC detection: No tasks created for %d volumes (skipped: %d already EC, %d too small, %d filtered, %d not quiet, %d not full)",
-			totalVolumes, skippedAlreadyEC, skippedTooSmall, skippedCollectionFilter, skippedQuietTime, skippedFullness)
+		glog.V(1).Infof("EC detection: No tasks created for %d volumes (skipped: %d already EC, %d too small, %d filtered, %d not quiet, %d not full, %d remote, %d too few nodes)",
+			totalVolumes, skippedAlreadyEC, skippedTooSmall, skippedCollectionFilter, skippedQuietTime, skippedFullness, skippedRemote, skippedTooFewNodes)
 
 		// Show details for first few volumes
 		for i, metric := range metrics {
@@ -658,6 +680,25 @@ func (p *ecPlacementPlanner) buildCandidateSets(shardsNeeded int) [][]*placement
 
 // planECDestinations plans the destinations for erasure coding operation.
 // dataShards/parityShards are parameters so callers can drive non-10+4 ratios.
+// countTopologyNodes counts volume-server nodes in the active topology, used by
+// the min-node safety gate.
+func countTopologyNodes(at *topology.ActiveTopology) int {
+	if at == nil {
+		return 0
+	}
+	topo := at.GetTopologyInfo()
+	if topo == nil {
+		return 0
+	}
+	n := 0
+	for _, dc := range topo.DataCenterInfos {
+		for _, rack := range dc.RackInfos {
+			n += len(rack.DataNodeInfos)
+		}
+	}
+	return n
+}
+
 func planECDestinations(planner *ecPlacementPlanner, metric *types.VolumeHealthMetrics, ecConfig *Config, dataShards, parityShards int) (*topology.MultiDestinationPlan, error) {
 	if planner == nil || planner.activeTopology == nil {
 		return nil, fmt.Errorf("active topology not available for EC placement")