Kevin Veen-BirkenbachandClaude Opus 5
07d819cb41
rsync -a expands to -rlptgoD, so the backup stored neither extended attributes nor ACLs nor hardlinks while the restore side asks for -aHAX. A generation therefore never held what a restore wants back, and rsync faithfully mirrored a snapshot that has none. Reproduced end to end against a source carrying user.oc.* attributes and a hardlink: with the old argv, getfattr -d -R over the generation returns nothing and the hardlink splits into two inodes; with -aHAX both sides come back identical to the source. That is the shape in which an OpenCloud volume survived a disaster-recovery drill with its files intact and its storage metadata gone, so its posix driver skipped every space it could no longer read. restore/files.py had the same gap on this repository's own restore path. Two consequences worth knowing. The first backup after this re-copies every attribute-carrying volume once, because rsync declines to hardlink against a generation whose xattrs do not match. And a backup destination on a filesystem without xattr support now fails the run instead of producing a generation that cannot be restored. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
94 lines
2.7 KiB
Python
94 lines
2.7 KiB
Python
from __future__ import annotations
|
|
|
|
import json
|
|
import os
|
|
import pathlib
|
|
from dataclasses import dataclass, field
|
|
|
|
from baudolo.generation import FILES_DIR
|
|
|
|
from .shell import BackupError, execute_shell_command
|
|
|
|
|
|
@dataclass(frozen=True)
|
|
class Backing:
|
|
"""Where a docker volume actually keeps its data.
|
|
|
|
Args:
|
|
mountpoint: the path the daemon reports.
|
|
driver: the volume driver, ``local`` for the built-in one.
|
|
options: the driver options; a non-empty map means the mountpoint is a
|
|
mount target rather than the storage itself.
|
|
"""
|
|
|
|
mountpoint: str
|
|
driver: str = "local"
|
|
options: dict = field(default_factory=dict)
|
|
|
|
@property
|
|
def source(self) -> str:
|
|
return f"{self.mountpoint}/"
|
|
|
|
|
|
def inspect_backing(volume_name: str) -> Backing:
|
|
reported = execute_shell_command(
|
|
["docker", "volume", "inspect", "--format", "{{json .}}", volume_name]
|
|
)[0]
|
|
data = json.loads(reported)
|
|
return Backing(
|
|
data.get("Mountpoint") or "",
|
|
data.get("Driver") or "",
|
|
data.get("Options") or {},
|
|
)
|
|
|
|
|
|
def get_last_backup_dir(
|
|
versions_dir: str, volume_name: str, current_backup_dir: str
|
|
) -> str | None:
|
|
versions = sorted(os.listdir(versions_dir), reverse=True)
|
|
for version in versions:
|
|
candidate = f"{pathlib.Path(versions_dir) / version / volume_name / FILES_DIR}/"
|
|
if candidate != current_backup_dir and pathlib.Path(candidate).is_dir():
|
|
return candidate
|
|
return None
|
|
|
|
|
|
def backup_volume(
|
|
versions_dir: str,
|
|
volume_name: str,
|
|
volume_dir: str,
|
|
*,
|
|
authoritative: bool,
|
|
source: str,
|
|
) -> None:
|
|
"""Perform incremental file backup of a Docker volume.
|
|
|
|
Args:
|
|
authoritative: compare source and destination by content instead of by
|
|
size and whole-second mtime. Required on a pass whose destination was
|
|
already written from a live source, where a file can differ while
|
|
both attributes still agree.
|
|
source: directory to read from - the volume's mountpoint, or its path
|
|
inside a snapshot.
|
|
"""
|
|
dest = f"{pathlib.Path(volume_dir) / FILES_DIR}/"
|
|
pathlib.Path(dest).mkdir(parents=True, exist_ok=True)
|
|
|
|
last = get_last_backup_dir(versions_dir, volume_name, dest)
|
|
cmd = ["rsync", "-aHAX", "-P", "--no-D", "--delete", "--delete-excluded"]
|
|
if authoritative:
|
|
cmd.append("--checksum")
|
|
if last:
|
|
cmd.append(f"--link-dest={last}")
|
|
cmd += [source, dest]
|
|
|
|
try:
|
|
execute_shell_command(cmd)
|
|
except BackupError as e:
|
|
if "file has vanished" in str(e):
|
|
print(
|
|
"Warning: Some files vanished before transfer. Continuing.", flush=True
|
|
)
|
|
else:
|
|
raise
|