debian-mirror-gitlab/lib/backup/files.rb

148 lines
5 KiB
Ruby
Raw Normal View History

2018-12-05 23:21:45 +05:30
# frozen_string_literal: true
2015-11-26 14:37:03 +05:30
require 'open3'
2018-05-09 12:01:36 +05:30
require_relative 'helper'
2015-11-26 14:37:03 +05:30
module Backup
class Files
2018-05-09 12:01:36 +05:30
include Backup::Helper
2020-11-24 15:15:51 +05:30
DEFAULT_EXCLUDE = 'lost+found'
2015-11-26 14:37:03 +05:30
2020-11-24 15:15:51 +05:30
attr_reader :name, :app_files_dir, :backup_tarball, :excludes, :files_parent_dir
def initialize(name, app_files_dir, excludes: [])
2015-11-26 14:37:03 +05:30
@name = name
@app_files_dir = File.realpath(app_files_dir)
@files_parent_dir = File.realpath(File.join(@app_files_dir, '..'))
2017-08-17 22:00:37 +05:30
@backup_files_dir = File.join(Gitlab.config.backup.path, File.basename(@app_files_dir) )
2015-11-26 14:37:03 +05:30
@backup_tarball = File.join(Gitlab.config.backup.path, name + '.tar.gz')
2020-11-24 15:15:51 +05:30
@excludes = [DEFAULT_EXCLUDE].concat(excludes)
2015-11-26 14:37:03 +05:30
end
# Copy files from public/files to backup/files
def dump
FileUtils.mkdir_p(Gitlab.config.backup.path)
FileUtils.rm_f(backup_tarball)
2017-08-17 22:00:37 +05:30
if ENV['STRATEGY'] == 'copy'
2021-02-22 17:27:13 +05:30
cmd = [%w[rsync -a --delete], exclude_dirs(:rsync), %W[#{app_files_dir} #{Gitlab.config.backup.path}]].flatten
2017-08-17 22:00:37 +05:30
output, status = Gitlab::Popen.popen(cmd)
2021-02-22 17:27:13 +05:30
# Retry if rsync source files vanish
if status == 24
$stdout.puts "Warning: files vanished during rsync, retrying..."
output, status = Gitlab::Popen.popen(cmd)
end
2020-10-24 23:57:45 +05:30
unless status == 0
2017-08-17 22:00:37 +05:30
puts output
2018-11-08 19:23:39 +05:30
raise Backup::Error, 'Backup failed'
2017-08-17 22:00:37 +05:30
end
2021-01-29 00:20:46 +05:30
tar_cmd = [tar, exclude_dirs(:tar), %W[-C #{@backup_files_dir} -cf - .]].flatten
status_list, output = run_pipeline!([tar_cmd, gzip_cmd], out: [backup_tarball, 'w', 0600])
2017-08-17 22:00:37 +05:30
FileUtils.rm_rf(@backup_files_dir)
else
2021-01-29 00:20:46 +05:30
tar_cmd = [tar, exclude_dirs(:tar), %W[-C #{app_files_dir} -cf - .]].flatten
status_list, output = run_pipeline!([tar_cmd, gzip_cmd], out: [backup_tarball, 'w', 0600])
end
unless pipeline_succeeded?(tar_status: status_list[0], gzip_status: status_list[1], output: output)
raise Backup::Error, "Backup operation failed: #{output}"
2017-08-17 22:00:37 +05:30
end
2015-11-26 14:37:03 +05:30
end
def restore
backup_existing_files_dir
2021-01-29 00:20:46 +05:30
cmd_list = [%w[gzip -cd], %W[#{tar} --unlink-first --recursive-unlink -C #{app_files_dir} -xf -]]
status_list, output = run_pipeline!(cmd_list, in: backup_tarball)
unless pipeline_succeeded?(gzip_status: status_list[0], tar_status: status_list[1], output: output)
raise Backup::Error, "Restore operation failed: #{output}"
end
2018-11-08 19:23:39 +05:30
end
def tar
if system(*%w[gtar --version], out: '/dev/null')
# It looks like we can get GNU tar by running 'gtar'
'gtar'
else
'tar'
end
2015-11-26 14:37:03 +05:30
end
def backup_existing_files_dir
2018-05-09 12:01:36 +05:30
timestamped_files_path = File.join(Gitlab.config.backup.path, "tmp", "#{name}.#{Time.now.to_i}")
2016-09-13 17:45:13 +05:30
if File.exist?(app_files_dir)
2018-05-09 12:01:36 +05:30
# Move all files in the existing repos directory except . and .. to
# repositories.old.<timestamp> directory
FileUtils.mkdir_p(timestamped_files_path, mode: 0700)
files = Dir.glob(File.join(app_files_dir, "*"), File::FNM_DOTMATCH) - [File.join(app_files_dir, "."), File.join(app_files_dir, "..")]
begin
FileUtils.mv(files, timestamped_files_path)
rescue Errno::EACCES
access_denied_error(app_files_dir)
2018-10-15 14:42:47 +05:30
rescue Errno::EBUSY
resource_busy_error(app_files_dir)
2018-05-09 12:01:36 +05:30
end
2015-11-26 14:37:03 +05:30
end
end
2017-08-17 22:00:37 +05:30
def run_pipeline!(cmd_list, options = {})
2019-03-02 22:35:43 +05:30
err_r, err_w = IO.pipe
options[:err] = err_w
2021-01-29 00:20:46 +05:30
status_list = Open3.pipeline(*cmd_list, options)
2019-03-02 22:35:43 +05:30
err_w.close
2021-01-29 00:20:46 +05:30
[status_list, err_r.read]
end
def noncritical_warning?(warning)
noncritical_warnings = [
/^g?tar: \.: Cannot mkdir: No such file or directory$/
]
noncritical_warnings.map { |w| warning =~ w }.any?
end
def pipeline_succeeded?(tar_status:, gzip_status:, output:)
return false unless gzip_status&.success?
tar_status&.success? || tar_ignore_non_success?(tar_status.exitstatus, output)
end
def tar_ignore_non_success?(exitstatus, output)
# tar can exit with nonzero code:
# 1 - if some files changed (i.e. a CI job is currently writes to log)
# 2 - if it cannot create `.` directory (see issue https://gitlab.com/gitlab-org/gitlab/-/issues/22442)
# http://www.gnu.org/software/tar/manual/html_section/tar_19.html#Synopsis
# so check tar status 1 or stderr output against some non-critical warnings
if exitstatus == 1
$stdout.puts "Ignoring tar exit status 1 'Some files differ': #{output}"
return true
end
# allow tar to fail with other non-success status if output contain non-critical warning
if noncritical_warning?(output)
$stdout.puts "Ignoring non-success exit status #{exitstatus} due to output of non-critical warning(s): #{output}"
return true
end
false
2015-11-26 14:37:03 +05:30
end
2020-11-24 15:15:51 +05:30
def exclude_dirs(fmt)
excludes.map do |s|
if s == DEFAULT_EXCLUDE
'--exclude=' + s
elsif fmt == :rsync
2021-03-11 19:13:27 +05:30
'--exclude=/' + File.join(File.basename(app_files_dir), s)
2020-11-24 15:15:51 +05:30
elsif fmt == :tar
'--exclude=./' + s
end
end
end
2015-11-26 14:37:03 +05:30
end
end