From caaa3de4ff62a117ba286cd1ec19b54a6c84f393 Mon Sep 17 00:00:00 2001
From: Miek Gieben <miek@improbable.io>
Date: Wed, 12 Aug 2015 15:44:42 +0100
Subject: [PATCH] Make HashMod use MD5 instead of FNV

MD5 will will distribute the inputs more uniformly over the output
space than FNV; leading to more evenly balanced load when using HashMod.
---
 retrieval/relabel.go      | 18 ++++++++++++++----
 retrieval/relabel_test.go |  2 +-
 2 files changed, 15 insertions(+), 5 deletions(-)

diff --git a/retrieval/relabel.go b/retrieval/relabel.go
index 0b2785f58..9d06bc0d2 100644
--- a/retrieval/relabel.go
+++ b/retrieval/relabel.go
@@ -1,8 +1,8 @@
 package retrieval
 
 import (
+	"crypto/md5"
 	"fmt"
-	"hash/fnv"
 	"strings"
 
 	clientmodel "github.com/prometheus/client_golang/model"
@@ -58,12 +58,22 @@ func relabel(labels clientmodel.LabelSet, cfg *config.RelabelConfig) (clientmode
 			labels[cfg.TargetLabel] = clientmodel.LabelValue(res)
 		}
 	case config.RelabelHashMod:
-		hasher := fnv.New64a()
-		hasher.Write([]byte(val))
-		mod := hasher.Sum64() % cfg.Modulus
+		mod := sum64(md5.Sum([]byte(val))) % cfg.Modulus
 		labels[cfg.TargetLabel] = clientmodel.LabelValue(fmt.Sprintf("%d", mod))
 	default:
 		panic(fmt.Errorf("retrieval.relabel: unknown relabel action type %q", cfg.Action))
 	}
 	return labels, nil
 }
+
+// sum64 sums the md5 hash to an uint64.
+func sum64(hash [md5.Size]byte) uint64 {
+	var s uint64
+
+	for i, b := range hash {
+		shift := uint64((md5.Size - i - 1) * 8)
+
+		s |= uint64(b) << shift
+	}
+	return s
+}
diff --git a/retrieval/relabel_test.go b/retrieval/relabel_test.go
index 455067c83..a788c03d4 100644
--- a/retrieval/relabel_test.go
+++ b/retrieval/relabel_test.go
@@ -170,7 +170,7 @@ func TestRelabel(t *testing.T) {
 				"a": "foo",
 				"b": "bar",
 				"c": "baz",
-				"d": "58",
+				"d": "976",
 			},
 		},
 	}