Add support for zstd compression (#249)

Co-authored-by: Michal Middleton <jafa81@gmail.com>

README.md

@@ -148,13 +148,22 @@ You can populate below template according to your requirements and use it as you
 
 # BACKUP_CRON_EXPRESSION="0 2 * * *"
 
-# The name of the backup file including the `.tar.gz` extension.
+# The compression algorithm used in conjunction with tar.
+# Valid options are: "gz" (Gzip) and "zst" (Zstd).
+# Note that the selection affects the file extension.
+
+# BACKUP_COMPRESSION="gz"
+
+# The name of the backup file including the extension.
 # Format verbs will be replaced as in `strftime`. Omitting them
 # will result in the same filename for every backup run, which means previous
-# versions will be overwritten on subsequent runs. The default results
+# versions will be overwritten on subsequent runs.
-# in filenames like `backup-2021-08-29T04-00-00.tar.gz`.
+# Extension can be defined literally or via "{{ .Extension }}" template,
+# in which case it will become either "tar.gz" or "tar.zst" (depending
+# on your BACKUP_COMPRESSION setting).
+# The default results in filenames like: `backup-2021-08-29T04-00-00.tar.gz`.
 
-# BACKUP_FILENAME="backup-%Y-%m-%dT%H-%M-%S.tar.gz"
+# BACKUP_FILENAME="backup-%Y-%m-%dT%H-%M-%S.{{ .Extension }}"
 
 # Setting BACKUP_FILENAME_EXPAND to true allows for environment variable
 # placeholders in BACKUP_FILENAME, BACKUP_LATEST_SYMLINK and in

cmd/backup/archive.go

@@ -15,9 +15,11 @@ import (
 	"path"
 	"path/filepath"
 	"strings"
+
+	"github.com/klauspost/compress/zstd"
 )
 
-func createArchive(files []string, inputFilePath, outputFilePath string) error {
+func createArchive(files []string, inputFilePath, outputFilePath string, compression string) error {
 	inputFilePath = stripTrailingSlashes(inputFilePath)
 	inputFilePath, outputFilePath, err := makeAbsolute(inputFilePath, outputFilePath)
 	if err != nil {
@@ -27,7 +29,7 @@ func createArchive(files []string, inputFilePath, outputFilePath string) error {
 		return fmt.Errorf("createArchive: error creating output file path: %w", err)
 	}
 
-	if err := compress(files, outputFilePath, filepath.Dir(inputFilePath)); err != nil {
+	if err := compress(files, outputFilePath, filepath.Dir(inputFilePath), compression); err != nil {
 		return fmt.Errorf("createArchive: error creating archive: %w", err)
 	}
 
@@ -51,18 +53,30 @@ func makeAbsolute(inputFilePath, outputFilePath string) (string, string, error)
 	return inputFilePath, outputFilePath, err
 }
 
-func compress(paths []string, outFilePath, subPath string) error {
+func compress(paths []string, outFilePath, subPath string, algo string) error {
 	file, err := os.Create(outFilePath)
+	var compressWriter io.WriteCloser
 	if err != nil {
 		return fmt.Errorf("compress: error creating out file: %w", err)
 	}
 
 	prefix := path.Dir(outFilePath)
-	gzipWriter := gzip.NewWriter(file)
+	switch algo {
-	tarWriter := tar.NewWriter(gzipWriter)
+	case "gz":
+		compressWriter = gzip.NewWriter(file)
+	case "zst":
+		compressWriter, err = zstd.NewWriter(file)
+		if err != nil {
+			return fmt.Errorf("compress: zstd error: %w", err)
+		}
+	default:
+		return fmt.Errorf("compress: unsupported compression algorithm: %s", algo)
+	}
+
+	tarWriter := tar.NewWriter(compressWriter)
 
 	for _, p := range paths {
-		if err := writeTarGz(p, tarWriter, prefix); err != nil {
+		if err := writeTarball(p, tarWriter, prefix); err != nil {
 			return fmt.Errorf("compress: error writing %s to archive: %w", p, err)
 		}
 	}
@@ -72,9 +86,9 @@ func compress(paths []string, outFilePath, subPath string) error {
 		return fmt.Errorf("compress: error closing tar writer: %w", err)
 	}
 
-	err = gzipWriter.Close()
+	err = compressWriter.Close()
 	if err != nil {
-		return fmt.Errorf("compress: error closing gzip writer: %w", err)
+		return fmt.Errorf("compress: error closing compression writer: %w", err)
 	}
 
 	err = file.Close()
@@ -85,10 +99,10 @@ func compress(paths []string, outFilePath, subPath string) error {
 	return nil
 }
 
-func writeTarGz(path string, tarWriter *tar.Writer, prefix string) error {
+func writeTarball(path string, tarWriter *tar.Writer, prefix string) error {
 	fileInfo, err := os.Lstat(path)
 	if err != nil {
-		return fmt.Errorf("writeTarGz: error getting file infor for %s: %w", path, err)
+		return fmt.Errorf("writeTarball: error getting file infor for %s: %w", path, err)
 	}
 
 	if fileInfo.Mode()&os.ModeSocket == os.ModeSocket {
@@ -99,19 +113,19 @@ func writeTarGz(path string, tarWriter *tar.Writer, prefix string) error {
 	if fileInfo.Mode()&os.ModeSymlink == os.ModeSymlink {
 		var err error
 		if link, err = os.Readlink(path); err != nil {
-			return fmt.Errorf("writeTarGz: error resolving symlink %s: %w", path, err)
+			return fmt.Errorf("writeTarball: error resolving symlink %s: %w", path, err)
 		}
 	}
 
 	header, err := tar.FileInfoHeader(fileInfo, link)
 	if err != nil {
-		return fmt.Errorf("writeTarGz: error getting file info header: %w", err)
+		return fmt.Errorf("writeTarball: error getting file info header: %w", err)
 	}
 	header.Name = strings.TrimPrefix(path, prefix)
 
 	err = tarWriter.WriteHeader(header)
 	if err != nil {
-		return fmt.Errorf("writeTarGz: error writing file info header: %w", err)
+		return fmt.Errorf("writeTarball: error writing file info header: %w", err)
 	}
 
 	if !fileInfo.Mode().IsRegular() {
@@ -120,13 +134,13 @@ func writeTarGz(path string, tarWriter *tar.Writer, prefix string) error {
 
 	file, err := os.Open(path)
 	if err != nil {
-		return fmt.Errorf("writeTarGz: error opening %s: %w", path, err)
+		return fmt.Errorf("writeTarball: error opening %s: %w", path, err)
 	}
 	defer file.Close()
 
 	_, err = io.Copy(tarWriter, file)
 	if err != nil {
-		return fmt.Errorf("writeTarGz: error copying %s to tar writer: %w", path, err)
+		return fmt.Errorf("writeTarball: error copying %s to tar writer: %w", path, err)
 	}
 
 	return nil

cmd/backup/config.go

@@ -16,59 +16,60 @@ import (
 // Config holds all configuration values that are expected to be set
 // by users.
 type Config struct {
-	AwsS3BucketName               string        `split_words:"true"`
+	AwsS3BucketName               string          `split_words:"true"`
-	AwsS3Path                     string        `split_words:"true"`
+	AwsS3Path                     string          `split_words:"true"`
-	AwsEndpoint                   string        `split_words:"true" default:"s3.amazonaws.com"`
+	AwsEndpoint                   string          `split_words:"true" default:"s3.amazonaws.com"`
-	AwsEndpointProto              string        `split_words:"true" default:"https"`
+	AwsEndpointProto              string          `split_words:"true" default:"https"`
-	AwsEndpointInsecure           bool          `split_words:"true"`
+	AwsEndpointInsecure           bool            `split_words:"true"`
-	AwsEndpointCACert             CertDecoder   `envconfig:"AWS_ENDPOINT_CA_CERT"`
+	AwsEndpointCACert             CertDecoder     `envconfig:"AWS_ENDPOINT_CA_CERT"`
-	AwsStorageClass               string        `split_words:"true"`
+	AwsStorageClass               string          `split_words:"true"`
-	AwsAccessKeyID                string        `envconfig:"AWS_ACCESS_KEY_ID"`
+	AwsAccessKeyID                string          `envconfig:"AWS_ACCESS_KEY_ID"`
-	AwsAccessKeyIDFile            string        `envconfig:"AWS_ACCESS_KEY_ID_FILE"`
+	AwsAccessKeyIDFile            string          `envconfig:"AWS_ACCESS_KEY_ID_FILE"`
-	AwsSecretAccessKey            string        `split_words:"true"`
+	AwsSecretAccessKey            string          `split_words:"true"`
-	AwsSecretAccessKeyFile        string        `split_words:"true"`
+	AwsSecretAccessKeyFile        string          `split_words:"true"`
-	AwsIamRoleEndpoint            string        `split_words:"true"`
+	AwsIamRoleEndpoint            string          `split_words:"true"`
-	AwsPartSize                   int64         `split_words:"true"`
+	AwsPartSize                   int64           `split_words:"true"`
-	BackupSources                 string        `split_words:"true" default:"/backup"`
+	BackupCompression             CompressionType `split_words:"true" default:"gz"`
-	BackupFilename                string        `split_words:"true" default:"backup-%Y-%m-%dT%H-%M-%S.tar.gz"`
+	BackupSources                 string          `split_words:"true" default:"/backup"`
-	BackupFilenameExpand          bool          `split_words:"true"`
+	BackupFilename                string          `split_words:"true" default:"backup-%Y-%m-%dT%H-%M-%S.{{ .Extension }}"`
-	BackupLatestSymlink           string        `split_words:"true"`
+	BackupFilenameExpand          bool            `split_words:"true"`
-	BackupArchive                 string        `split_words:"true" default:"/archive"`
+	BackupLatestSymlink           string          `split_words:"true"`
-	BackupRetentionDays           int32         `split_words:"true" default:"-1"`
+	BackupArchive                 string          `split_words:"true" default:"/archive"`
-	BackupPruningLeeway           time.Duration `split_words:"true" default:"1m"`
+	BackupRetentionDays           int32           `split_words:"true" default:"-1"`
-	BackupPruningPrefix           string        `split_words:"true"`
+	BackupPruningLeeway           time.Duration   `split_words:"true" default:"1m"`
-	BackupStopContainerLabel      string        `split_words:"true" default:"true"`
+	BackupPruningPrefix           string          `split_words:"true"`
-	BackupFromSnapshot            bool          `split_words:"true"`
+	BackupStopContainerLabel      string          `split_words:"true" default:"true"`
-	BackupExcludeRegexp           RegexpDecoder `split_words:"true"`
+	BackupFromSnapshot            bool            `split_words:"true"`
-	GpgPassphrase                 string        `split_words:"true"`
+	BackupExcludeRegexp           RegexpDecoder   `split_words:"true"`
-	NotificationURLs              []string      `envconfig:"NOTIFICATION_URLS"`
+	GpgPassphrase                 string          `split_words:"true"`
-	NotificationLevel             string        `split_words:"true" default:"error"`
+	NotificationURLs              []string        `envconfig:"NOTIFICATION_URLS"`
-	EmailNotificationRecipient    string        `split_words:"true"`
+	NotificationLevel             string          `split_words:"true" default:"error"`
-	EmailNotificationSender       string        `split_words:"true" default:"noreply@nohost"`
+	EmailNotificationRecipient    string          `split_words:"true"`
-	EmailSMTPHost                 string        `envconfig:"EMAIL_SMTP_HOST"`
+	EmailNotificationSender       string          `split_words:"true" default:"noreply@nohost"`
-	EmailSMTPPort                 int           `envconfig:"EMAIL_SMTP_PORT" default:"587"`
+	EmailSMTPHost                 string          `envconfig:"EMAIL_SMTP_HOST"`
-	EmailSMTPUsername             string        `envconfig:"EMAIL_SMTP_USERNAME"`
+	EmailSMTPPort                 int             `envconfig:"EMAIL_SMTP_PORT" default:"587"`
-	EmailSMTPPassword             string        `envconfig:"EMAIL_SMTP_PASSWORD"`
+	EmailSMTPUsername             string          `envconfig:"EMAIL_SMTP_USERNAME"`
-	WebdavUrl                     string        `split_words:"true"`
+	EmailSMTPPassword             string          `envconfig:"EMAIL_SMTP_PASSWORD"`
-	WebdavUrlInsecure             bool          `split_words:"true"`
+	WebdavUrl                     string          `split_words:"true"`
-	WebdavPath                    string        `split_words:"true" default:"/"`
+	WebdavUrlInsecure             bool            `split_words:"true"`
-	WebdavUsername                string        `split_words:"true"`
+	WebdavPath                    string          `split_words:"true" default:"/"`
-	WebdavPassword                string        `split_words:"true"`
+	WebdavUsername                string          `split_words:"true"`
-	SSHHostName                   string        `split_words:"true"`
+	WebdavPassword                string          `split_words:"true"`
-	SSHPort                       string        `split_words:"true" default:"22"`
+	SSHHostName                   string          `split_words:"true"`
-	SSHUser                       string        `split_words:"true"`
+	SSHPort                       string          `split_words:"true" default:"22"`
-	SSHPassword                   string        `split_words:"true"`
+	SSHUser                       string          `split_words:"true"`
-	SSHIdentityFile               string        `split_words:"true" default:"/root/.ssh/id_rsa"`
+	SSHPassword                   string          `split_words:"true"`
-	SSHIdentityPassphrase         string        `split_words:"true"`
+	SSHIdentityFile               string          `split_words:"true" default:"/root/.ssh/id_rsa"`
-	SSHRemotePath                 string        `split_words:"true"`
+	SSHIdentityPassphrase         string          `split_words:"true"`
-	ExecLabel                     string        `split_words:"true"`
+	SSHRemotePath                 string          `split_words:"true"`
-	ExecForwardOutput             bool          `split_words:"true"`
+	ExecLabel                     string          `split_words:"true"`
-	LockTimeout                   time.Duration `split_words:"true" default:"60m"`
+	ExecForwardOutput             bool            `split_words:"true"`
-	AzureStorageAccountName       string        `split_words:"true"`
+	LockTimeout                   time.Duration   `split_words:"true" default:"60m"`
-	AzureStoragePrimaryAccountKey string        `split_words:"true"`
+	AzureStorageAccountName       string          `split_words:"true"`
-	AzureStorageContainerName     string        `split_words:"true"`
+	AzureStoragePrimaryAccountKey string          `split_words:"true"`
-	AzureStoragePath              string        `split_words:"true"`
+	AzureStorageContainerName     string          `split_words:"true"`
-	AzureStorageEndpoint          string        `split_words:"true" default:"https://{{ .AccountName }}.blob.core.windows.net/"`
+	AzureStoragePath              string          `split_words:"true"`
+	AzureStorageEndpoint          string          `split_words:"true" default:"https://{{ .AccountName }}.blob.core.windows.net/"`
 }
 
 func (c *Config) resolveSecret(envVar string, secretPath string) (string, error) {
@@ -82,6 +83,22 @@ func (c *Config) resolveSecret(envVar string, secretPath string) (string, error)
 	return string(data), nil
 }
 
+type CompressionType string
+
+func (c *CompressionType) Decode(v string) error {
+	switch v {
+	case "gz", "zst":
+		*c = CompressionType(v)
+		return nil
+	default:
+		return fmt.Errorf("config: error decoding compression type %s", v)
+	}
+}
+
+func (c *CompressionType) String() string {
+	return string(*c)
+}
+
 type CertDecoder struct {
 	Cert *x509.Certificate
 }

cmd/backup/script.go

@@ -4,6 +4,7 @@
 package main
 
 import (
+	"bytes"
 	"context"
 	"errors"
 	"fmt"
@@ -89,6 +90,20 @@ func newScript() (*script, error) {
 	}
 
 	s.file = path.Join("/tmp", s.c.BackupFilename)
+
+	tmplFileName, tErr := template.New("extension").Parse(s.file)
+	if tErr != nil {
+		return nil, fmt.Errorf("newScript: unable to parse backup file extension template: %w", tErr)
+	}
+
+	var bf bytes.Buffer
+	if tErr := tmplFileName.Execute(&bf, map[string]string{
+		"Extension": fmt.Sprintf("tar.%s", s.c.BackupCompression),
+	}); tErr != nil {
+		return nil, fmt.Errorf("newScript: error executing backup file extension template: %w", tErr)
+	}
+	s.file = bf.String()
+
 	if s.c.BackupFilenameExpand {
 		s.file = os.ExpandEnv(s.file)
 		s.c.BackupLatestSymlink = os.ExpandEnv(s.c.BackupLatestSymlink)
@@ -454,7 +469,7 @@ func (s *script) createArchive() error {
 		return fmt.Errorf("createArchive: error walking filesystem tree: %w", err)
 	}
 
-	if err := createArchive(filesEligibleForBackup, backupSources, tarFile); err != nil {
+	if err := createArchive(filesEligibleForBackup, backupSources, tarFile, s.c.BackupCompression.String()); err != nil {
 		return fmt.Errorf("createArchive: error compressing backup folder: %w", err)
 	}
 

go.mod

@@ -10,6 +10,7 @@ require (
 	github.com/docker/docker v24.0.5+incompatible
 	github.com/gofrs/flock v0.8.1
 	github.com/kelseyhightower/envconfig v1.4.0
+	github.com/klauspost/compress v1.16.7
 	github.com/leekchan/timeutil v0.0.0-20150802142658-28917288c48d
 	github.com/minio/minio-go/v7 v7.0.61
 	github.com/otiai10/copy v1.11.0
@@ -33,7 +34,6 @@ require (
 	github.com/golang-jwt/jwt/v4 v4.5.0 // indirect
 	github.com/google/uuid v1.3.0 // indirect
 	github.com/json-iterator/go v1.1.12 // indirect
-	github.com/klauspost/compress v1.16.7 // indirect
 	github.com/klauspost/cpuid/v2 v2.2.5 // indirect
 	github.com/kr/fs v0.1.0 // indirect
 	github.com/kylelemons/godebug v1.1.0 // indirect

test/cli-zstd/run.sh

@@ -0,0 +1,66 @@
+#!/bin/sh
+
+set -e
+
+cd $(dirname $0)
+. ../util.sh
+current_test=$(basename $(pwd))
+
+docker network create test_network
+docker volume create backup_data
+docker volume create app_data
+# This volume is created to test whether empty directories are handled
+# correctly. It is not supposed to hold any data.
+docker volume create empty_data
+
+docker run -d \
+  --name minio \
+  --network test_network \
+  --env MINIO_ROOT_USER=test \
+  --env MINIO_ROOT_PASSWORD=test \
+  --env MINIO_ACCESS_KEY=test \
+  --env MINIO_SECRET_KEY=GMusLtUmILge2by+z890kQ \
+  -v backup_data:/data \
+  minio/minio:RELEASE.2020-08-04T23-10-51Z server /data
+
+docker exec minio mkdir -p /data/backup
+
+docker run -d \
+  --name offen \
+  --network test_network \
+  -v app_data:/var/opt/offen/ \
+  offen/offen:latest
+
+sleep 10
+
+docker run --rm \
+  --network test_network \
+  -v app_data:/backup/app_data \
+  -v empty_data:/backup/empty_data \
+  -v /var/run/docker.sock:/var/run/docker.sock \
+  --env AWS_ACCESS_KEY_ID=test \
+  --env AWS_SECRET_ACCESS_KEY=GMusLtUmILge2by+z890kQ \
+  --env AWS_ENDPOINT=minio:9000 \
+  --env AWS_ENDPOINT_PROTO=http \
+  --env AWS_S3_BUCKET_NAME=backup \
+  --env BACKUP_COMPRESSION=zst \
+  --env BACKUP_FILENAME='test.{{ .Extension }}' \
+  --env "BACKUP_FROM_SNAPSHOT=true" \
+  --entrypoint backup \
+  offen/docker-volume-backup:${TEST_VERSION:-canary}
+
+# Have to install tar and zstd on Alpine because the plain image comes with very
+# basic tar from busybox and it does not seem to support zstd
+docker run --rm \
+  -v backup_data:/data alpine \
+  ash -c 'apk add --no-cache zstd tar && tar -xvf /data/backup/test.tar.zst --zstd && test -f /backup/app_data/offen.db && test -d /backup/empty_data'
+
+pass "Found relevant files in untared remote backup."
+
+# This test does not stop containers during backup. This is happening on
+# purpose in order to cover this setup as well.
+expect_running_containers "2"
+
+docker rm $(docker stop minio offen)
+docker volume rm backup_data app_data
+docker network rm test_network