X-Git-Url: https://git.arvados.org/arvados.git/blobdiff_plain/6fe8e52020d421797306e5c6536afbcee761510a..e99f026d040c6020dfcc51c6d988cf18d325a530:/services/keep-balance/block_state.go

diff --git a/services/keep-balance/block_state.go b/services/keep-balance/block_state.go
index 22e89c019a..e30b4ff794 100644
--- a/services/keep-balance/block_state.go
+++ b/services/keep-balance/block_state.go
@@ -7,7 +7,7 @@ package main
 import (
 	"sync"
 
-	"git.curoverse.com/arvados.git/sdk/go/arvados"
+	"git.arvados.org/arvados.git/sdk/go/arvados"
 )
 
 // Replica is a file on disk (or object in an S3 bucket, or blob in an
@@ -23,6 +23,8 @@ type Replica struct {
 // replicas actually stored (according to the keepstore indexes we
 // know about).
 type BlockState struct {
+	Refs     map[string]bool // pdh => true (only tracked when len(Replicas)==0)
+	RefCount int
 	Replicas []Replica
 	Desired  map[string]int
 	// TODO: Support combinations of classes ("private + durable")
@@ -39,9 +41,22 @@ var defaultClasses = []string{"default"}
 
 func (bs *BlockState) addReplica(r Replica) {
 	bs.Replicas = append(bs.Replicas, r)
+	// Free up memory wasted by tracking PDHs that will never be
+	// reported (see comment in increaseDesired)
+	bs.Refs = nil
 }
 
-func (bs *BlockState) increaseDesired(classes []string, n int) {
+func (bs *BlockState) increaseDesired(pdh string, classes []string, n int) {
+	if pdh != "" && len(bs.Replicas) == 0 {
+		// Note we only track PDHs if there's a possibility
+		// that we will report the list of referring PDHs,
+		// i.e., if we haven't yet seen a replica.
+		if bs.Refs == nil {
+			bs.Refs = map[string]bool{}
+		}
+		bs.Refs[pdh] = true
+	}
+	bs.RefCount++
 	if len(classes) == 0 {
 		classes = defaultClasses
 	}
@@ -107,11 +122,64 @@ func (bsm *BlockStateMap) AddReplicas(mnt *KeepMount, idx []arvados.KeepServiceI
 
 // IncreaseDesired updates the map to indicate the desired replication
 // for the given blocks in the given storage class is at least n.
-func (bsm *BlockStateMap) IncreaseDesired(classes []string, n int, blocks []arvados.SizedDigest) {
+//
+// If pdh is non-empty, it will be tracked and reported in the "lost
+// blocks" report.
+func (bsm *BlockStateMap) IncreaseDesired(pdh string, classes []string, n int, blocks []arvados.SizedDigest) {
 	bsm.mutex.Lock()
 	defer bsm.mutex.Unlock()
 
 	for _, blkid := range blocks {
-		bsm.get(blkid).increaseDesired(classes, n)
+		bsm.get(blkid).increaseDesired(pdh, classes, n)
+	}
+}
+
+// GetConfirmedReplication returns the replication level of the given
+// blocks, considering only the specified storage classes.
+//
+// If len(classes)==0, returns the replication level without regard to
+// storage classes.
+//
+// Safe to call concurrently with other calls to GetCurrent, but not
+// with different BlockStateMap methods.
+func (bsm *BlockStateMap) GetConfirmedReplication(blkids []arvados.SizedDigest, classes []string) int {
+	defaultClasses := map[string]bool{"default": true}
+	min := 0
+	for _, blkid := range blkids {
+		total := 0
+		perclass := make(map[string]int, len(classes))
+		for _, c := range classes {
+			perclass[c] = 0
+		}
+		for _, r := range bsm.get(blkid).Replicas {
+			total += r.KeepMount.Replication
+			mntclasses := r.KeepMount.StorageClasses
+			if len(mntclasses) == 0 {
+				mntclasses = defaultClasses
+			}
+			for c := range mntclasses {
+				n, ok := perclass[c]
+				if !ok {
+					// Don't care about this storage class
+					continue
+				}
+				perclass[c] = n + r.KeepMount.Replication
+			}
+		}
+		if total == 0 {
+			return 0
+		}
+		for _, n := range perclass {
+			if n == 0 {
+				return 0
+			}
+			if n < min || min == 0 {
+				min = n
+			}
+		}
+		if len(perclass) == 0 && (min == 0 || min > total) {
+			min = total
+		}
 	}
+	return min
 }