diff --git a/api.go b/api.go
index ed0a925a6..c745b2a0b 100644
--- a/api.go
+++ b/api.go
@@ -697,7 +697,7 @@ func (api *API) ImportValue(ctx context.Context, req internal.ImportValueRequest
 
 // MaxShards returns the maximum shard number for each index in a map.
 func (api *API) MaxShards(ctx context.Context) map[string]uint64 {
-	return api.holder.MaxShards()
+	return api.holder.maxShards()
 }
 
 // StatsWithTags returns an instance of whatever implementation of StatsClient
diff --git a/cluster.go b/cluster.go
index a5108a2f9..a3f5aa775 100644
--- a/cluster.go
+++ b/cluster.go
@@ -1150,7 +1150,7 @@ func (c *cluster) generateResizeJobByAction(nodeAction nodeAction) (*resizeJob,
 			Node:          EncodeNode(toCluster.unprotectedNodeByID(id)),
 			Coordinator:   EncodeNode(c.coordinatorNode()),
 			Sources:       sources,
-			Schema:        c.holder.EncodeSchema(), // Include the schema to ensure it's in sync on the receiving node.
+			Schema:        c.holder.encodeSchema(), // Include the schema to ensure it's in sync on the receiving node.
 			ClusterStatus: c.Status(),
 		}
 		j.Instructions = append(j.Instructions, instr)
@@ -1205,7 +1205,7 @@ func (c *cluster) followResizeInstruction(instr *internal.ResizeInstruction) err
 
 			// Sync the schema received in the resize instruction.
 			c.logger.Printf("Holder ApplySchema")
-			if err := c.holder.ApplySchema(instr.Schema); err != nil {
+			if err := c.holder.applySchema(instr.Schema); err != nil {
 				return errors.Wrap(err, "applying schema")
 			}
 
diff --git a/holder.go b/holder.go
index bad001787..8cf830a84 100644
--- a/holder.go
+++ b/holder.go
@@ -36,8 +36,8 @@ const (
 	// defaultCacheFlushInterval is the default value for Fragment.CacheFlushInterval.
 	defaultCacheFlushInterval = 1 * time.Minute
 
-	// FileLimit is the maximum open file limit (ulimit -n) to automatically set.
-	FileLimit = 262144 // (512^2)
+	// fileLimit is the maximum open file limit (ulimit -n) to automatically set.
+	fileLimit = 262144 // (512^2)
 )
 
 // Holder represents a container for indexes.
@@ -50,7 +50,7 @@ type Holder struct {
 	// opened channel is closed once Open() completes.
 	opened chan struct{}
 
-	Broadcaster broadcaster
+	broadcaster broadcaster
 
 	NewAttrStore func(string) AttrStore
 
@@ -65,7 +65,7 @@ type Holder struct {
 	Path string
 
 	// The interval at which the cached row ids are persisted to disk.
-	CacheFlushInterval time.Duration
+	cacheFlushInterval time.Duration
 
 	Logger Logger
 }
@@ -78,12 +78,12 @@ func NewHolder() *Holder {
 
 		opened: make(chan struct{}),
 
-		Broadcaster: NopBroadcaster,
+		broadcaster: NopBroadcaster,
 		Stats:       NopStatsClient,
 
 		NewAttrStore: newNopAttrStore,
 
-		CacheFlushInterval: defaultCacheFlushInterval,
+		cacheFlushInterval: defaultCacheFlushInterval,
 
 		Logger: NopLogger,
 	}
@@ -200,8 +200,8 @@ func (h *Holder) HasData() (bool, error) {
 	return false, nil
 }
 
-// MaxShards returns MaxShard map for all indexes.
-func (h *Holder) MaxShards() map[string]uint64 {
+// maxShards returns MaxShard map for all indexes.
+func (h *Holder) maxShards() map[string]uint64 {
 	a := make(map[string]uint64)
 	for _, index := range h.Indexes() {
 		a[index.Name()] = index.MaxShard()
@@ -229,8 +229,8 @@ func (h *Holder) Schema() []*IndexInfo {
 	return a
 }
 
-// ApplySchema applies an internal Schema to Holder.
-func (h *Holder) ApplySchema(schema *internal.Schema) error {
+// applySchema applies an internal Schema to Holder.
+func (h *Holder) applySchema(schema *internal.Schema) error {
 	// Create indexes that don't exist.
 	for _, index := range schema.Indexes {
 		opt := IndexOptions{}
@@ -257,15 +257,15 @@ func (h *Holder) ApplySchema(schema *internal.Schema) error {
 	return nil
 }
 
-// EncodeMaxShards creates and internal representation of max shards.
-func (h *Holder) EncodeMaxShards() *internal.MaxShards {
+// encodeMaxShards creates and internal representation of max shards.
+func (h *Holder) encodeMaxShards() *internal.MaxShards {
 	return &internal.MaxShards{
-		Standard: h.MaxShards(),
+		Standard: h.maxShards(),
 	}
 }
 
-// EncodeSchema creates an internal representation of schema.
-func (h *Holder) EncodeSchema() *internal.Schema {
+// encodeSchema creates an internal representation of schema.
+func (h *Holder) encodeSchema() *internal.Schema {
 	return &internal.Schema{
 		Indexes: EncodeIndexes(h.Indexes()),
 	}
@@ -360,7 +360,7 @@ func (h *Holder) newIndex(path, name string) (*Index, error) {
 	}
 	index.Logger = h.Logger
 	index.Stats = h.Stats.WithTags(fmt.Sprintf("index:%s", index.Name()))
-	index.broadcaster = h.Broadcaster
+	index.broadcaster = h.broadcaster
 	index.NewAttrStore = h.NewAttrStore
 	index.columnAttrStore = h.NewAttrStore(filepath.Join(index.path, ".data"))
 	return index, nil
@@ -423,7 +423,7 @@ func (h *Holder) fragment(index, field, view string, shard uint64) *Fragment {
 // monitorCacheFlush periodically flushes all fragment caches sequentially.
 // This is run in a goroutine.
 func (h *Holder) monitorCacheFlush() {
-	ticker := time.NewTicker(h.CacheFlushInterval)
+	ticker := time.NewTicker(h.cacheFlushInterval)
 	defer ticker.Stop()
 
 	for {
@@ -476,11 +476,11 @@ func (h *Holder) setFileLimit() {
 		return
 	}
 	// If the soft limit is lower than the FileLimit constant, we will try to change it.
-	if oldLimit.Cur < FileLimit {
-		newLimit.Cur = FileLimit
+	if oldLimit.Cur < fileLimit {
+		newLimit.Cur = fileLimit
 		// If the hard limit is not high enough, we will try to change it too.
-		if oldLimit.Max < FileLimit {
-			newLimit.Max = FileLimit
+		if oldLimit.Max < fileLimit {
+			newLimit.Max = fileLimit
 		} else {
 			newLimit.Max = oldLimit.Max
 		}
@@ -508,8 +508,8 @@ func (h *Holder) setFileLimit() {
 		if err := syscall.Getrlimit(syscall.RLIMIT_NOFILE, oldLimit); err != nil {
 			h.Logger.Printf("ERROR checking open file limit: %s", err)
 		} else {
-			if oldLimit.Cur < FileLimit {
-				h.Logger.Printf("WARNING: Tried to set open file limit to %d, but it is %d. You may consider running \"sudo ulimit -n %d\" before starting Pilosa to avoid \"too many open files\" error. See https://www.pilosa.com/docs/administration/#open-file-limits for more information.", FileLimit, oldLimit.Cur, FileLimit)
+			if oldLimit.Cur < fileLimit {
+				h.Logger.Printf("WARNING: Tried to set open file limit to %d, but it is %d. You may consider running \"sudo ulimit -n %d\" before starting Pilosa to avoid \"too many open files\" error. See https://www.pilosa.com/docs/administration/#open-file-limits for more information.", fileLimit, oldLimit.Cur, fileLimit)
 			}
 		}
 	}
diff --git a/server.go b/server.go
index 0e03365ae..93a5dac36 100644
--- a/server.go
+++ b/server.go
@@ -298,7 +298,7 @@ func NewServer(opts ...ServerOption) (*Server, error) {
 	s.executor.MaxWritesPerRequest = s.maxWritesPerRequest
 	s.cluster.broadcaster = s
 	s.cluster.maxWritesPerRequest = s.maxWritesPerRequest
-	s.holder.Broadcaster = s
+	s.holder.broadcaster = s
 
 	err = s.cluster.setup()
 	if err != nil {
@@ -572,8 +572,8 @@ func (s *Server) LocalStatus() (proto.Message, error) {
 
 	ns := internal.NodeStatus{
 		Node:      EncodeNode(s.cluster.Node),
-		MaxShards: s.holder.EncodeMaxShards(),
-		Schema:    s.holder.EncodeSchema(),
+		MaxShards: s.holder.encodeMaxShards(),
+		Schema:    s.holder.encodeSchema(),
 	}
 
 	return &ns, nil
@@ -606,12 +606,12 @@ func (s *Server) mergeRemoteStatus(ns *internal.NodeStatus) error {
 	}
 
 	// Sync schema.
-	if err := s.holder.ApplySchema(ns.Schema); err != nil {
+	if err := s.holder.applySchema(ns.Schema); err != nil {
 		return errors.Wrap(err, "applying schema")
 	}
 
 	// Sync maxShards.
-	oldmaxshards := s.holder.MaxShards()
+	oldmaxshards := s.holder.maxShards()
 	for index, newMax := range ns.MaxShards.Standard {
 		localIndex := s.holder.Index(index)
 		// if we don't know about an index locally, log an error because
diff --git a/utils_internal_test.go b/utils_internal_test.go
index f7309961c..ca7bd1fa7 100644
--- a/utils_internal_test.go
+++ b/utils_internal_test.go
@@ -360,7 +360,7 @@ func (t *ClusterCluster) FollowResizeInstruction(instr *internal.ResizeInstructi
 		destCluster := t.clusterByID(instrNode.ID)
 
 		// Sync the schema received in the resize instruction.
-		if err := destCluster.holder.ApplySchema(instr.Schema); err != nil {
+		if err := destCluster.holder.applySchema(instr.Schema); err != nil {
 			return err
 		}