Initial commit

This commit is contained in:
dev
2025-02-27 21:53:53 +08:00
commit 815e55e4c0
1291 changed files with 185445 additions and 0 deletions

37
hf3fs_utils/README.md Normal file
View File

@@ -0,0 +1,37 @@
# hf3fs_cli
build:
```bash
python3 setup_hf3fs_utils.py bdist_wheel
```
usage:
```.bash
$ hf3fs_cli rmtree --help
Usage: hf3fs_cli rmtree [OPTIONS] [DIR_PATHS]...
Move a directory tree to the trash and set an expiration time, it will be automatically deleted after expiration
Example:
hf3fs_cli rmtree <path/to/remove> --expire <expire_time>
- Use --expire [1h|3h|8h|1d|3d|7d] to specify the expiration time, the directory will be deleted after expiration.
- Before expiration, you can restore the directory from the trash using `hf3fs_cli mv <trash_path> <target_path>`.
- If you need to free up space immediately, you can use `hf3fs_cli rmtree <trash_path>` to delete the data in the trash immediately, this operation cannot be undone!
- Use `ls /path/to/hf3fs/trash` to view the trash.
Options:
--expire [1h|3h|8h|1d|3d|7d] Expiration time, contents in the trash will be automatically deleted after expiration
-y, --yes Skip confirmation prompt and delete immediately
--help Show this message and exit.
$ hf3fs_cli mv --help
Usage: hf3fs_cli mv [OPTIONS] OLD_PATH NEW_PATH
Move files, supports moving files between different mount points within the same 3FS
Options:
--help Show this message and exit.
```
If you want to use `rmtree` command, the administrator needs to create a trash directory for each user at `/{3fs_mountpoint}/trash/{user_name}`. The cleanup of the trash directory is handled by the `trash_cleaner`. For instructions on how to use it, please refer to `src/client/trash_cleaner/`.

0
hf3fs_utils/__init__.py Normal file
View File

192
hf3fs_utils/cli.py Normal file
View File

@@ -0,0 +1,192 @@
import errno
import click
import os
import sys
import stat
from typing import Optional, List
from hf3fs_utils.fs import is_relative_to, FileSystem
from hf3fs_utils.trash import TRASH_CONFIGS, Trash
MOUNTPOINT = os.environ.get("HF3FS_CLI_MOUNTPOINT", None)
def get_filesystem(path: str) -> FileSystem:
mountpoint = None
if MOUNTPOINT is not None:
mountpoint = os.path.abspath(MOUNTPOINT)
else:
path = os.path.realpath(path)
parts = path.split(os.sep)
for i in range(1, 4):
p = os.sep.join(parts[:i])
if os.path.exists(os.path.join(p, "3fs-virt")):
mountpoint = p
break
if not mountpoint:
abort(f"{path} is not on 3FS")
return FileSystem(mountpoint)
def abs_path(path: str) -> str:
if ".." in path.split(os.path.sep):
abort(f"Path {path} contains '..', which is not supported yet")
normpath = os.path.normpath(path)
# If the user calls rmtree path/symlink, it should delete the symlink instead of the path it points to
# For dir paths, take the realpath, but keep the filename as is
dir = os.path.dirname(normpath)
filename = os.path.basename(normpath)
return os.path.join(os.path.realpath(dir), filename)
def abort(msg):
click.echo(click.style(msg, fg="red"), err=True)
sys.exit(1)
@click.group()
def cli():
"""
3FS command-line tool
"""
@cli.command()
@click.argument("old_path", type=click.Path(exists=True))
@click.argument("new_path", type=click.Path())
def mv(old_path: str, new_path: str):
"""
Move files, supports moving files between different mount points within the same 3FS
"""
try:
old_path = abs_path(old_path)
new_path = abs_path(new_path)
try:
new_st = os.stat(new_path, follow_symlinks=True)
# new_path exists, should be a directory
if not stat.S_ISDIR(new_st.st_mode):
raise FileExistsError(errno.EEXIST, os.strerror(errno.EEXIST), new_path)
# move to new_path/filename
new_path = os.path.join(new_path, os.path.basename(old_path))
except FileNotFoundError:
pass
fs = get_filesystem(old_path)
fs.rename(old_path, new_path)
click.echo(f"Move successful: {old_path} -> {new_path}")
except AssertionError:
raise
except Exception as ex:
abort(f"Move failed: {ex}")
class ExpireType(click.ParamType):
def get_metavar(self, param) -> str:
return "[1h|3h|8h|1d|3d|7d]"
def convert(self, value, param, ctx):
norm_value = value
if norm_value.endswith("hour"):
norm_value = norm_value.replace("hour", "h")
elif norm_value.endswith("hours"):
norm_value = norm_value.replace("hours", "h")
elif norm_value.endswith("day"):
norm_value = norm_value.replace("day", "d")
elif norm_value.endswith("days"):
norm_value = norm_value.replace("days", "d")
if norm_value not in TRASH_CONFIGS.keys():
self.fail(f"{value} is invalid, valid options are {self.get_metavar()}", param, ctx)
else:
return norm_value
@cli.command()
@click.argument("dir_paths", type=click.Path(exists=True), nargs=-1)
@click.option(
"--expire",
type=ExpireType(),
help="Expiration time, contents in the trash will be automatically deleted after expiration",
)
@click.option("-y", "--yes", is_flag=True, default=False, help="Skip confirmation prompt and delete immediately")
def rmtree(dir_paths: List[str], expire: Optional[str], yes: bool):
"""
Move a directory tree to the trash and set an expiration time, it will be automatically deleted after expiration
\b
Example:
hf3fs_cli rmtree <path/to/remove> --expire <expire_time>
\b
- Use --expire [1h|3h|8h|1d|3d|7d] to specify the expiration time, the directory will be deleted after expiration.
- Before expiration, you can restore the directory from the trash using `hf3fs_cli mv <trash_path> <target_path>`.
- If you need to free up space immediately, you can use `hf3fs_cli rmtree <trash_path>` to delete the data in the trash immediately, this operation cannot be undone!
- Use `ls /hf3fs/{cluster}/trash` to view the trash.
"""
if not dir_paths:
abort(f"Please provide the directory path to delete")
first_path = abs_path(dir_paths[0])
fs = get_filesystem(first_path)
fs_trash = Trash(fs)
clean_trash = is_relative_to(first_path, fs_trash.trash_path)
if not clean_trash:
if not expire:
abort(f"Use --expire [1h|3h|8h|1d|3d|7d] to specify the expiration time")
elif expire:
abort(f"{first_path} is already in the trash")
trash_cfg = TRASH_CONFIGS[expire] if not clean_trash else None
dir_paths = [abs_path(p) for p in dir_paths]
for dir_path in dir_paths:
if is_relative_to(dir_path, fs_trash.trash_path) != clean_trash:
if clean_trash:
abort(f"{dir_path} is not in the trash")
else:
abort(f"{dir_path} is already in the trash")
if clean_trash:
if len(dir_paths) != 1:
msg = (
f"Immediately delete the following paths:\n"
+ "\n".join([f"- {p}" for p in dir_paths])
+ "\nThis operation cannot be undone"
)
else:
msg = f"Immediately delete {dir_path}, this operation cannot be undone"
else:
if len(dir_paths) != 1:
msg = (
f"Move the following paths to the trash:\n"
+ "\n".join([f"- {p}" for p in dir_paths])
+ f"\nThey will be automatically deleted after {expire}"
)
else:
msg = f"Move {dir_path} to the trash, it will be automatically deleted after {expire}"
if not yes:
assert click.confirm(msg, abort=True)
for dir_path in dir_paths:
try:
if clean_trash:
fs.remove(dir_path, recursive=True)
click.echo(f"- Deleted {dir_path}")
else:
trash_path = fs_trash.move_to_trash(dir_path, trash_cfg)
click.echo(f"- Trash path: {trash_path}")
except AssertionError:
raise
except Exception as ex:
abort(f"Failed to delete {dir_path}: {ex}")
if not clean_trash:
click.echo(
"- Before expiration, you can use 'hf3fs_cli mv <trash_path> <target_path>' to restore, "
"or use 'hf3fs_cli rmtree <trash_path>' to delete immediately and free up space"
)
if __name__ == "__main__":
cli()

234
hf3fs_utils/fs.py Normal file
View File

@@ -0,0 +1,234 @@
import os
import fcntl
import errno
import struct
import stat
import sys
import pathlib
from typing import Tuple
def is_relative_to(path1, path2) -> bool:
try:
pathlib.PurePath(path1).relative_to(path2)
return True
except:
return False
class FileSystem:
HF3FS_IOCTL_MAGIC_CMD = 0x80046802
HF3FS_IOCTL_MAGIC_NUM = 0x8F3F5FFF
HF3FS_IOCTL_VERSION_CMD = 0x80046803
HF3FS_IOCTL_RENAME_CMD = 0x4218680E
HF3FS_IOCTL_RENAME_BUFFER_SIZE = 536
HF3FS_IOCTL_REMOVE_CMD = 0x4110680F
HF3FS_IOCTL_REMOVE_BUFFER_SIZE = 272
def __init__(self, mountpoint: str) -> None:
self.mountpoint = os.path.realpath(mountpoint)
self.virt_path = os.path.join(self.mountpoint, "3fs-virt")
# Check if the mount point is a directory
if not os.path.exists(self.mountpoint):
raise FileNotFoundError(
errno.ENOENT, os.strerror(errno.ENOENT), self.mountpoint
)
if not os.path.isdir(self.mountpoint):
raise NotADirectoryError(
errno.ENOTDIR, os.strerror(errno.ENOTDIR), self.mountpoint
)
virt_fd = None
try:
# Check the 3fs-virt directory
virt_fd = os.open(self.virt_path, os.O_RDONLY | os.O_DIRECTORY)
virt_st = os.fstat(virt_fd)
if not stat.S_ISDIR(virt_st.st_mode):
raise NotADirectoryError(
errno.ENOTDIR, os.strerror(errno.ENOTDIR), self.virt_path
)
self.st_dev = virt_st.st_dev
# Check the magic number
buffer = bytearray(4)
try:
self._ioctl(virt_fd, FileSystem.HF3FS_IOCTL_MAGIC_CMD, buffer)
except OSError:
raise RuntimeError(f"{self.mountpoint} is not a 3FS mount point")
magic_number = struct.unpack("I", buffer)[0]
expected_magic_number = FileSystem.HF3FS_IOCTL_MAGIC_NUM
if magic_number != expected_magic_number:
raise RuntimeError(
f"{self.mountpoint} is not a 3FS mount point, "
f"magic number {magic_number:x} != {expected_magic_number:x}"
)
# Check if the required ioctl is supported
ioctl_version = -1
try:
buffer = bytearray(4)
self._ioctl(virt_fd, FileSystem.HF3FS_IOCTL_VERSION_CMD, buffer)
ioctl_version = int.from_bytes(buffer, sys.byteorder, signed=False)
except OSError:
pass
assert ioctl_version >= 1
finally:
if virt_fd is not None:
os.close(virt_fd)
def _check_user(self):
if os.geteuid() == 0 or os.getegid() == 0:
raise RuntimeError(f"root user not allowed")
def _encode_filename(self, name: str) -> bytes:
assert name and os.sep not in name, name
name_bytes = name.encode("utf8")
if len(name_bytes) > 255:
raise OSError(errno.ENAMETOOLONG, os.strerror(errno.ENAMETOOLONG), name)
return name_bytes
def opendir(self, dir_path: str) -> Tuple[int, os.stat_result]:
dir_fd = os.open(dir_path, os.O_DIRECTORY | os.O_RDONLY)
try:
try:
dir_st = os.fstat(dir_fd)
except OSError as ex:
ex.filename = dir_path
raise
if not stat.S_ISDIR(dir_st.st_mode):
raise NotADirectoryError(
errno.ENOTDIR, os.strerror(errno.ENOTDIR), dir_path
)
if dir_st.st_dev != self.st_dev:
raise RuntimeError(f"{dir_path} is not under the 3FS mount point {self.mountpoint}")
if dir_st.st_ino & 0xF000000000000000:
raise RuntimeError(f"{dir_path} is a virtual path")
return dir_fd, dir_st
except:
os.close(dir_fd)
raise
def split_path(self, path: str) -> Tuple[int, os.stat_result, str]:
filename = os.path.basename(path)
if not filename:
raise RuntimeError(f"{path} has no filename")
if filename in [".", ".."]:
raise RuntimeError(f"{path} filename is {filename}")
if len(filename.encode("utf8")) > 255:
raise OSError(errno.ENAMETOOLONG, os.strerror(errno.ENAMETOOLONG), path)
dir = os.path.dirname(path) or "."
dir_fd, dir_st = self.opendir(dir)
return dir_fd, dir_st, filename
def rename(self, old_path: str, new_path: str) -> None:
self._check_user()
if is_relative_to(
os.path.realpath(new_path), os.path.join(self.mountpoint, "trash")
):
raise RuntimeError(f"{new_path} is in the trash")
old_dir_fd = None
new_dir_fd = None
try:
old_dir_fd, old_dir_st, old_filename = self.split_path(old_path)
new_dir_fd, new_dir_st, new_filename = self.split_path(new_path)
try:
old_st = os.stat(old_filename, dir_fd=old_dir_fd, follow_symlinks=False)
if stat.S_ISLNK(old_st.st_mode):
raise RuntimeError(f"{old_path} is symlink")
except OSError as ex:
ex.filename = old_path
raise
try:
os.stat(new_filename, dir_fd=new_dir_fd, follow_symlinks=False)
raise FileExistsError(errno.EEXIST, os.strerror(errno.EEXIST), new_path)
except FileNotFoundError:
pass
except OSError as ex:
ex.filename = new_path
raise
try:
self._rename_ioctl(
old_dir_fd,
old_dir_st.st_ino,
old_filename,
new_dir_st.st_ino,
new_filename,
False,
)
except OSError as ex:
ex.filename = old_path
ex.filename2 = new_path
raise
finally:
if old_dir_fd is not None:
os.close(old_dir_fd)
if new_dir_fd is not None:
os.close(new_dir_fd)
def _rename_ioctl(
self,
old_dir_fd: int,
old_dir_ino: int,
old_filename: str,
new_dir_ino: int,
new_filename: str,
move_to_trash: bool,
) -> None:
assert old_filename and not os.path.sep in old_filename, old_filename
assert new_filename and not os.path.sep in new_filename, new_filename
cmd = FileSystem.HF3FS_IOCTL_RENAME_CMD
buffer = struct.pack(
"N256sN256s?",
old_dir_ino,
self._encode_filename(old_filename),
new_dir_ino,
self._encode_filename(new_filename),
move_to_trash,
).ljust(FileSystem.HF3FS_IOCTL_RENAME_BUFFER_SIZE)
self._ioctl(old_dir_fd, cmd, buffer)
def remove(self, path: str, recursive: bool) -> None:
dir_fd = None
try:
dir_fd, dir_st, filename = self.split_path(path)
st = os.stat(filename, dir_fd=dir_fd, follow_symlinks=False)
if stat.S_ISLNK(st.st_mode):
raise RuntimeError(f"{path} is symlink")
if stat.S_ISDIR(st.st_mode) and recursive:
# The user must be the owner of the directory and have rwx permissions
imode = stat.S_IMODE(st.st_mode)
if st.st_uid != os.geteuid() or (imode & 0o700) != 0o700:
raise PermissionError(errno.EPERM, os.strerror(errno.EPERM), path)
try:
self._remove_ioctl(dir_fd, dir_st.st_ino, filename, recursive)
except OSError as ex:
ex.filename = path
raise ex
finally:
if dir_fd is not None:
os.close(dir_fd)
def _remove_ioctl(
self, parent_fd: int, parent_ino: int, filename: str, recursive: bool
) -> None:
assert filename and os.sep not in filename, filename
cmd = FileSystem.HF3FS_IOCTL_REMOVE_CMD
buffer = struct.pack(
"N256s?", parent_ino, self._encode_filename(filename), recursive
).ljust(FileSystem.HF3FS_IOCTL_REMOVE_BUFFER_SIZE)
self._ioctl(parent_fd, cmd, buffer)
def _ioctl(self, fd: int, cmd: int, buffer):
self._check_user()
return fcntl.ioctl(fd, cmd, buffer)

5
hf3fs_utils/hf3fs_cli Normal file
View File

@@ -0,0 +1,5 @@
#!/usr/bin/env python
from hf3fs_utils import cli
if __name__ == "__main__":
cli.cli(max_content_width=120)

176
hf3fs_utils/trash.py Normal file
View File

@@ -0,0 +1,176 @@
import os
import dataclasses
import pwd
import stat
import errno
import time
from typing import Optional
from datetime import datetime, timedelta, timezone
from . import fs
UTC8_TZ = timezone(timedelta(hours=8))
DATE_FORMAT = "%Y%m%d_%H%M"
BASE_TIMESTAMP = int(datetime(year=1980, month=1, day=1, tzinfo=UTC8_TZ).timestamp())
def format_date(t: datetime) -> str:
assert t.tzinfo
return t.astimezone(tz=UTC8_TZ).strftime(DATE_FORMAT)
def parse_date(t: str) -> datetime:
return datetime.strptime(t, DATE_FORMAT).replace(tzinfo=UTC8_TZ)
def get_timestamp_us() -> int:
timestamp_seconds = time.time()
return int(timestamp_seconds * 1_000_000)
@dataclasses.dataclass
class TrashConfig:
name: str
expire: timedelta
time_slice: timedelta
def __post_init__(self):
assert self.name and "-" not in self.name, f"invalid name {self.name}"
assert self.expire >= timedelta(minutes=1), self.expire
assert self.time_slice >= timedelta(minutes=1), self.time_slice
assert self.time_slice < self.expire, (self.time_slice, self.expire)
def current_dir(self) -> str:
base_timestamp = BASE_TIMESTAMP
current_timestamp = int(datetime.now(tz=UTC8_TZ).timestamp())
assert current_timestamp > base_timestamp, current_timestamp
time_slice_seconds = int(self.time_slice.total_seconds())
expire_seconds = int(self.expire.total_seconds())
assert time_slice_seconds and expire_seconds, repr(self)
start_timestamp = (
(current_timestamp - base_timestamp) // time_slice_seconds
) * time_slice_seconds + base_timestamp
end_timestamp = start_timestamp + expire_seconds + time_slice_seconds
start_datetime = datetime.fromtimestamp(start_timestamp, tz=UTC8_TZ)
end_datetime = datetime.fromtimestamp(end_timestamp, tz=UTC8_TZ)
return f"{self.name}-{format_date(start_datetime)}-{format_date(end_datetime)}"
TRASH_CONFIGS = {
"1h": TrashConfig("1h", timedelta(hours=1), timedelta(minutes=10)),
"3h": TrashConfig("3h", timedelta(hours=3), timedelta(minutes=30)),
"8h": TrashConfig("8h", timedelta(hours=8), timedelta(minutes=30)),
"1d": TrashConfig("1d", timedelta(days=1), timedelta(hours=1)),
"3d": TrashConfig("3d", timedelta(days=3), timedelta(days=1)),
"7d": TrashConfig("7d", timedelta(days=7), timedelta(days=1)),
}
class Trash:
def __init__(
self,
filesystem: fs.FileSystem,
user: Optional[int] = None,
user_name: Optional[str] = None,
) -> None:
if user is None:
user = os.geteuid()
assert isinstance(user, int), user
if user_name is None:
user_name = pwd.getpwuid(user).pw_name
if user == 0:
raise RuntimeError(f"hf3fs trash does not support root user")
# Check if the trash directory is mounted
trash = os.path.join(filesystem.mountpoint, "trash")
if not os.path.exists(trash):
raise FileNotFoundError(errno.ENOENT, os.strerror(errno.ENOENT), trash)
# Check if the user's trash directory exists
user_trash = os.path.join(filesystem.mountpoint, "trash", user_name)
if not os.path.exists(user_trash):
raise FileNotFoundError(errno.ENOENT, os.strerror(errno.ENOENT), user_trash)
user_trash_fd, user_trash_st = filesystem.opendir(user_trash)
os.close(user_trash_fd)
assert stat.S_ISDIR(user_trash_st.st_mode)
if user_trash_st.st_uid != user:
raise RuntimeError(
f"Trash directory {user_trash}, owner {user_trash_st.st_uid} != {user}"
)
self.filesystem = filesystem
self.user = user
self.user_name = user_name
self.trash_path = trash
self.user_trash_path = user_trash
def _check_user(self):
euid = os.geteuid()
if euid != self.user:
raise RuntimeError(f"euid {euid} != trash owner {self.user}")
def move_to_trash(
self,
path: str,
config: TrashConfig,
trash_name: Optional[str] = None,
append_timestamp_if_exists: bool = True,
) -> str:
self._check_user()
assert isinstance(config, TrashConfig), f"invalid trash config {config}"
dir_fd = None
trash_dir_fd = None
try:
dir_fd, dir_st, filename = self.filesystem.split_path(path)
try:
st = os.stat(filename, dir_fd=dir_fd, follow_symlinks=False)
except OSError as ex:
ex.filename = path
raise
if stat.S_ISDIR(st.st_mode):
# The user must be the owner of the directory and have rwx permissions.
imode = stat.S_IMODE(st.st_mode)
if st.st_uid != os.geteuid() or (imode & 0o700) != 0o700:
raise PermissionError(errno.EPERM, os.strerror(errno.EPERM), path)
trash_dir = os.path.join(self.user_trash_path, config.current_dir())
try:
os.mkdir(trash_dir, 0o755)
except FileExistsError:
pass
trash_dir_fd, trash_dir_st = self.filesystem.opendir(trash_dir)
trash_name = trash_name or filename
current_trash_name = trash_name
retry = 0
while True:
retry += 1
try:
self.filesystem._rename_ioctl(
dir_fd,
dir_st.st_ino,
filename,
trash_dir_st.st_ino,
current_trash_name,
True,
)
return os.path.join(trash_dir, current_trash_name)
except OSError as ex:
if (
ex.errno in (errno.ENOTDIR, errno.EEXIST, errno.ENOTEMPTY)
and append_timestamp_if_exists
and retry < 10
):
current_trash_name = f"{trash_name[0:200]}.{get_timestamp_us()}"
else:
raise
finally:
if dir_fd is not None:
os.close(dir_fd)
if trash_dir_fd is not None:
os.close(trash_dir_fd)