diff --git a/checkpoint/checkpoint.go b/checkpoint/checkpoint.go
index f2c3bfacdab..2d0d108a590 100644
--- a/checkpoint/checkpoint.go
+++ b/checkpoint/checkpoint.go
@@ -22,6 +22,7 @@ package checkpoint
 
 import (
 	"fmt"
+	"io"
 	"io/ioutil"
 	"os"
 	"path/filepath"
@@ -32,6 +33,7 @@ import (
 	"gopkg.in/yaml.v2"
 
 	"github.com/elastic/beats/libbeat/logp"
+	"github.com/elastic/beats/libbeat/paths"
 )
 
 // Checkpoint persists event log state information to disk.
@@ -87,6 +89,11 @@ func NewCheckpoint(file string, maxUpdates int, interval time.Duration) (*Checkp
 		save:          make(chan JournalState, 1),
 	}
 
+	err := c.findRegistryFile()
+	if err != nil {
+		return nil, fmt.Errorf("error locating the proper registry file: %+v", err)
+	}
+
 	// Minimum batch size.
 	if c.maxUpdates < 1 {
 		c.maxUpdates = 1
@@ -120,6 +127,53 @@ func NewCheckpoint(file string, maxUpdates int, interval time.Duration) (*Checkp
 	return c, nil
 }
 
+// Previously the registry file was written to the root folder. It was fixed on
+// 7.x but not on 6.x. Thus, migration is needed, so users avoid losing state info.
+func (c *Checkpoint) findRegistryFile() error {
+	migratedPath := paths.Resolve(paths.Data, c.file)
+
+	fs, err := os.Stat(c.file)
+	if os.IsNotExist(err) {
+		c.file = migratedPath
+		return nil
+	} else if err != nil {
+		return fmt.Errorf("error accessing previous registry file: %+v", err)
+	}
+
+	f, err := os.Open(c.file)
+	if err != nil {
+		return err
+	}
+	defer f.Close()
+
+	target, err := os.OpenFile(migratedPath, os.O_WRONLY|os.O_CREATE|os.O_TRUNC, fs.Mode())
+	if err != nil {
+		return err
+	}
+	defer target.Close()
+
+	if _, err := io.Copy(target, f); err != nil {
+		return err
+	}
+
+	err = target.Sync()
+	if err != nil {
+		return fmt.Errorf("error while syncing new registry file to disk: %+v", err)
+	}
+
+	c.file = migratedPath
+
+	p := filepath.Dir(migratedPath)
+	pf, err := os.Open(p)
+	if err != nil {
+		return nil
+	}
+	defer pf.Close()
+	pf.Sync()
+
+	return nil
+}
+
 // run is worker loop that reads incoming state information from the save
 // channel and persists it when the number of changes reaches maxEvents or
 // the amount of time since the last disk write reaches flushInterval.