Coverage for bzfs_main/bzfs.py: 99%

1# 

2# Copyright 2024 Wolfgang Hoschek AT mac DOT com 

3# 

4# Licensed under the Apache License, Version 2.0 (the "License"); 

5# you may not use this file except in compliance with the License. 

6# You may obtain a copy of the License at 

7# 

8# http://www.apache.org/licenses/LICENSE-2.0 

9# 

10# Unless required by applicable law or agreed to in writing, software 

11# distributed under the License is distributed on an "AS IS" BASIS, 

12# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 

13# See the License for the specific language governing permissions and 

14# limitations under the License. 

15 

16# Inline script metadata conforming to https://packaging.python.org/specifications/inline-script-metadata 

17# /// script 

18# requires-python = ">=3.8" 

19# dependencies = [] 

20# /// 

21 

22""" 

23* Overview of the bzfs.py codebase: 

24* The codebase starts with docs, definition of input data and associated argument parsing into a "Params" class. 

25* All CLI option/parameter values are reachable from the "Params" class. 

26* Control flow starts in main(), far below ..., which kicks off a "Job". 

27* A Job runs one or more "tasks" via run_tasks(), each task replicating a separate dataset tree. 

28* The core replication algorithm is in run_task() and especially in replicate_datasets() and replicate_dataset(). 

29* The filter algorithms that apply include/exclude policies are in filter_datasets() and filter_snapshots(). 

30* The --create-src-snapshots-* and --delete-* and --compare-* and --monitor-* algorithms also start in run_task(). 

31* Consider using an IDE/editor that can open multiple windows for the same (long) file, such as PyCharm or Sublime Text, etc. 

32* The main retry logic is in run_with_retries() and clear_resumable_recv_state_if_necessary(). 

33* Progress reporting for use during `zfs send/recv` data transfers is in class ProgressReporter. 

34* Executing a CLI commmand on a local or remote host is in run_ssh_command(). 

35* Network connection management is in refresh_ssh_connection_if_necessary() and class ConnectionPool. 

36* Caching functionality can be found by searching for this regex: .*cach.* 

37* The parallel processing engine is in itr_ssh_cmd_parallel() and process_datasets_in_parallel_and_fault_tolerant(). 

38* README.md is mostly auto-generated from the ArgumentParser help texts as the source of "truth", via update_readme.sh. 

39Simply run that script whenever you change or add ArgumentParser help text. 

40""" 

41 

42import argparse 

43import ast 

44import bisect 

45import calendar 

46import collections 

47import concurrent 

48import contextlib 

49import copy 

50import dataclasses 

51import errno 

52import fcntl 

53import glob 

54import hashlib 

55import heapq 

56import itertools 

57import logging 

58import logging.config 

59import logging.handlers 

60import math 

61import operator 

62import os 

63import platform 

64import pwd 

65import random 

66import re 

67import selectors 

68import shlex 

69import shutil 

70import signal 

71import socket 

72import stat 

73import subprocess 

74import sys 

75import tempfile 

76import threading 

77import time 

78import types 

79from collections import defaultdict, deque, Counter 

80from concurrent.futures import ThreadPoolExecutor, Future, FIRST_COMPLETED 

81from dataclasses import dataclass, field 

82from datetime import datetime, timedelta, timezone, tzinfo 

83from logging import Logger 

84from os import stat as os_stat, utime as os_utime 

85from os.path import exists as os_path_exists, join as os_path_join 

86from pathlib import Path 

87from subprocess import CalledProcessError, DEVNULL, PIPE 

88from typing import ( 

89 Any, 

90 Callable, 

91 DefaultDict, 

92 Deque, 

93 Dict, 

94 Final, 

95 FrozenSet, 

96 Generator, 

97 Generic, 

98 ItemsView, 

99 Iterable, 

100 Iterator, 

101 List, 

102 Literal, 

103 NamedTuple, 

104 NoReturn, 

105 Optional, 

106 Protocol, 

107 Sequence, 

108 Set, 

109 TextIO, 

110 Tuple, 

111 Type, 

112 TypeVar, 

113 Union, 

114 cast, 

115) 

116 

117from bzfs_main.utils import cut 

118 

119# constants: 

120__version__ = "1.12.0-dev" 

121prog_name = "bzfs" 

122prog_author = "Wolfgang Hoschek" 

123die_status = 3 

124critical_status = 2 

125warning_status = 1 

126still_running_status = 4 

127min_python_version = (3, 8) 

128if sys.version_info < min_python_version: 

129 print(f"ERROR: {prog_name} requires Python version >= {'.'.join(map(str, min_python_version))}!") 

130 sys.exit(die_status) 

131exclude_dataset_regexes_default = r"(.*/)?[Tt][Ee]?[Mm][Pp][-_]?[0-9]*" # skip tmp datasets by default 

132create_src_snapshots_prefix_dflt = prog_name + "_" 

133create_src_snapshots_suffix_dflt = "_adhoc" 

134time_threshold_secs = 1.1 # 1 second ZFS creation time resolution + NTP clock skew is typically < 10ms 

135disable_prg = "-" 

136env_var_prefix = prog_name + "_" 

137pv_file_thread_separator = "_" 

138dummy_dataset = "dummy" 

139zfs_version_is_at_least_2_1_0 = "zfs>=2.1.0" 

140zfs_version_is_at_least_2_2_0 = "zfs>=2.2.0" 

141zfs_recv_groups = {"zfs_recv_o": "-o", "zfs_recv_x": "-x", "zfs_set": ""} 

142snapshot_regex_filter_names = frozenset({"include_snapshot_regex", "exclude_snapshot_regex"}) 

143snapshot_regex_filter_name = "snapshot_regex" 

144snapshot_filters_var = "snapshot_filters_var" 

145cmp_choices_items = ("src", "dst", "all") 

146inject_dst_pipe_fail_kbytes = 400 

147unixtime_infinity_secs = 2**64 # billions of years in the future and to be extra safe, larger than the largest ZFS GUID 

148year_with_four_digits_regex = re.compile(r"[1-9][0-9][0-9][0-9]") # regex for empty target shall not match non-empty target 

149log_stderr = (logging.INFO + logging.WARN) // 2 # custom log level is halfway in between 

150log_stdout = (log_stderr + logging.INFO) // 2 # custom log level is halfway in between 

151log_debug = logging.DEBUG 

152log_trace = logging.DEBUG // 2 # custom log level is halfway in between 

153SNAPSHOTS_CHANGED = "snapshots_changed" # See https://openzfs.github.io/openzfs-docs/man/7/zfsprops.7.html#snapshots_changed 

154BARRIER_CHAR = "~" 

155SHARED = "shared" 

156DEDICATED = "dedicated" 

157DONT_SKIP_DATASET = "" 

158SHELL_CHARS = '"' + "'`~!@#$%^&*()+={}[]|;<>?,\\" 

159 

160 

161def argument_parser() -> argparse.ArgumentParser: 

162 create_src_snapshots_plan_example1 = str({"test": {"": {"adhoc": 1}}}).replace(" ", "") 

163 create_src_snapshots_plan_example2 = str({"prod": {"us-west-1": {"hourly": 36, "daily": 31}}}).replace(" ", "") 

164 delete_dst_snapshots_except_plan_example1 = str( 

165 { 

166 "prod": { 

167 "onsite": { 

168 "secondly": 40, 

169 "minutely": 40, 

170 "hourly": 36, 

171 "daily": 31, 

172 "weekly": 12, 

173 "monthly": 18, 

174 "yearly": 5, 

175 } 

176 } 

177 } 

178 ).replace(" ", "") 

179 

180 # fmt: off 

181 parser = argparse.ArgumentParser( 

182 prog=prog_name, 

183 allow_abbrev=False, 

184 formatter_class=argparse.RawTextHelpFormatter, 

185 description=f""" 

186*{prog_name} is a backup command line tool that reliably replicates ZFS snapshots from a (local or remote) 

187source ZFS dataset (ZFS filesystem or ZFS volume) and its descendant datasets to a (local or remote) 

188destination ZFS dataset to make the destination dataset a recursively synchronized copy of the source dataset, 

189using zfs send/receive/rollback/destroy and ssh tunnel as directed. For example, {prog_name} can be used to 

190incrementally replicate all ZFS snapshots since the most recent common snapshot from source to destination, 

191in order to help protect against data loss or ransomware.* 

192 

193When run for the first time, {prog_name} replicates the dataset and all its snapshots from the source to the 

194destination. On subsequent runs, {prog_name} transfers only the data that has changed since the previous run, 

195i.e. it incrementally replicates to the destination all intermediate snapshots that have been created on 

196the source since the last run. Source ZFS snapshots older than the most recent common snapshot found on the 

197destination are auto-skipped. 

198 

199Unless {prog_name} is explicitly told to create snapshots on the source, it treats the source as read-only, 

200thus the source remains unmodified. With the --dryrun flag, {prog_name} also treats the destination as read-only. 

201In normal operation, {prog_name} treats the destination as append-only. Optional CLI flags are available to 

202delete destination snapshots and destination datasets as directed, for example to make the destination 

203identical to the source if the two have somehow diverged in unforeseen ways. This easily enables 

204(re)synchronizing the backup from the production state, as well as restoring the production state from 

205backup. 

206 

207In the spirit of rsync, {prog_name} supports a variety of powerful include/exclude filters that can be combined to 

208select which datasets, snapshots and properties to create, replicate, delete or compare. 

209 

210Typically, a `cron` job on the source host runs `{prog_name}` periodically to create new snapshots and prune outdated 

211snapshots on the source, whereas another `cron` job on the destination host runs `{prog_name}` periodically to prune 

212outdated destination snapshots. Yet another `cron` job runs `{prog_name}` periodically to replicate the recently created 

213snapshots from the source to the destination. The frequency of these periodic activities is typically every N milliseconds, 

214every second, minute, hour, day, week, month and/or year (or multiples thereof). 

215 

216All {prog_name} functions including snapshot creation, replication, deletion, monitoring, comparison, etc. happily work 

217with any snapshots in any format, even created or managed by third party ZFS snapshot management tools, including manual 

218zfs snapshot/destroy. All functions can also be used independently. That is, if you wish you can use {prog_name} just 

219for creating snapshots, or just for replicating, or just for deleting/pruning, or just for monitoring, or just for 

220comparing snapshot lists. 

221 

222The source 'pushes to' the destination whereas the destination 'pulls from' the source. {prog_name} is installed 

223and executed on the 'initiator' host which can be either the host that contains the source dataset (push mode), 

224or the destination dataset (pull mode), or both datasets (local mode, no network required, no ssh required), 

225or any third-party (even non-ZFS OSX) host as long as that host is able to SSH (via standard 'ssh' OpenSSH CLI) into 

226both the source and destination host (pull-push mode). In pull-push mode the source 'zfs send's the data stream 

227to the initiator which immediately pipes the stream (without storing anything locally) to the destination 

228host that 'zfs receive's it. Pull-push mode means that {prog_name} need not be installed or executed on either 

229source or destination host. Only the underlying 'zfs' CLI must be installed on both source and destination host. 

230{prog_name} can run as root or non-root user, in the latter case via a) sudo or b) when granted corresponding 

231ZFS permissions by administrators via 'zfs allow' delegation mechanism. 

232 

233{prog_name} is written in Python and continously runs a wide set of unit tests and integration tests to ensure 

234coverage and compatibility with old and new versions of ZFS on Linux, FreeBSD and Solaris, on all Python 

235versions >= 3.8 (including latest stable which is currently python-3.13). 

236 

237{prog_name} is a stand-alone program with zero required dependencies, akin to a 

238stand-alone shell script or binary executable. It is designed to be able to run in restricted barebones server 

239environments. No external Python packages are required; indeed no Python package management at all is required. 

240You can just symlink the program wherever you like, for example into /usr/local/bin or similar, and simply run it like 

241any stand-alone shell script or binary executable. 

242 

243{prog_name} automatically replicates the snapshots of multiple datasets in parallel for best performance. 

244Similarly, it quickly deletes (or monitors or compares) snapshots of multiple datasets in parallel. Atomic snapshots can be 

245created as frequently as every N milliseconds. 

246 

247Optionally, {prog_name} applies bandwidth rate-limiting and progress monitoring (via 'pv' CLI) during 'zfs 

248send/receive' data transfers. When run across the network, {prog_name} also transparently inserts lightweight 

249data compression (via 'zstd -1' CLI) and efficient data buffering (via 'mbuffer' CLI) into the pipeline 

250between network endpoints during 'zfs send/receive' network transfers. If one of these utilities is not 

251installed this is auto-detected, and the operation continues reliably without the corresponding auxiliary 

252feature. 

253 

254# Periodic Jobs with bzfs_jobrunner 

255 

256The software also ships with the [bzfs_jobrunner](README_bzfs_jobrunner.md) companion program, which is a convenience 

257wrapper around `{prog_name}` that simplifies efficient periodic ZFS snapshot creation, replication, pruning, and monitoring, 

258across N source hosts and M destination hosts, using a single shared [jobconfig](bzfs_tests/bzfs_job_example.py) script. 

259For example, this simplifies the deployment of an efficient geo-replicated backup service where each of the M destination 

260hosts is located in a separate geographic region and pulls replicas from (the same set of) N source hosts. It also 

261simplifies low latency replication from a primary to a secondary or to M read replicas, or backup to removable drives, etc. 

262 

263# Quickstart 

264 

265* Create adhoc atomic snapshots without a schedule: 

266 

267```$ {prog_name} tank1/foo/bar dummy --recursive --skip-replication --create-src-snapshots 

268--create-src-snapshots-plan "{create_src_snapshots_plan_example1}"``` 

269 

270```$ zfs list -t snapshot tank1/foo/bar 

271 

272tank1/foo/bar@test_2024-11-06_08:30:05_adhoc``` 

273 

274* Create periodic atomic snapshots on a schedule, every hour and every day, by launching this from a periodic `cron` job: 

275 

276```$ {prog_name} tank1/foo/bar dummy --recursive --skip-replication --create-src-snapshots 

277--create-src-snapshots-plan "{create_src_snapshots_plan_example2}"``` 

278 

279```$ zfs list -t snapshot tank1/foo/bar 

280 

281tank1/foo/bar@prod_us-west-1_2024-11-06_08:30:05_daily 

282 

283tank1/foo/bar@prod_us-west-1_2024-11-06_08:30:05_hourly``` 

284 

285Note: A periodic snapshot is created if it is due per the schedule indicated by its suffix (e.g. `_daily` or `_hourly` 

286or `_minutely` or `_2secondly` or `_100millisecondly`), or if the --create-src-snapshots-even-if-not-due flag is specified, 

287or if the most recent scheduled snapshot is somehow missing. In the latter case {prog_name} immediately creates a snapshot 

288(named with the current time, not backdated to the missed time), and then resumes the original schedule. If the suffix is 

289`_adhoc` or not a known period then a snapshot is considered non-periodic and is thus created immediately regardless of the 

290creation time of any existing snapshot. 

291 

292* Replication example in local mode (no network, no ssh), to replicate ZFS dataset tank1/foo/bar to tank2/boo/bar: 

293 

294```$ {prog_name} tank1/foo/bar tank2/boo/bar``` 

295 

296```$ zfs list -t snapshot tank1/foo/bar 

297 

298tank1/foo/bar@prod_us-west-1_2024-11-06_08:30:05_daily 

299 

300tank1/foo/bar@prod_us-west-1_2024-11-06_08:30:05_hourly``` 

301 

302```$ zfs list -t snapshot tank2/boo/bar 

303 

304tank2/boo/bar@prod_us-west-1_2024-11-06_08:30:05_daily 

305 

306tank2/boo/bar@prod_us-west-1_2024-11-06_08:30:05_hourly``` 

307 

308* Same example in pull mode: 

309 

310```$ {prog_name} root@host1.example.com:tank1/foo/bar tank2/boo/bar``` 

311 

312* Same example in push mode: 

313 

314```$ {prog_name} tank1/foo/bar root@host2.example.com:tank2/boo/bar``` 

315 

316* Same example in pull-push mode: 

317 

318```$ {prog_name} root@host1:tank1/foo/bar root@host2:tank2/boo/bar``` 

319 

320* Example in local mode (no network, no ssh) to recursively replicate ZFS dataset tank1/foo/bar and its descendant 

321datasets to tank2/boo/bar: 

322 

323```$ {prog_name} tank1/foo/bar tank2/boo/bar --recursive``` 

324 

325```$ zfs list -t snapshot -r tank1/foo/bar 

326 

327tank1/foo/bar@prod_us-west-1_2024-11-06_08:30:05_daily 

328 

329tank1/foo/bar@prod_us-west-1_2024-11-06_08:30:05_hourly 

330 

331tank1/foo/bar/baz@prod_us-west-1_2024-11-06_08:40:00_daily 

332 

333tank1/foo/bar/baz@prod_us-west-1_2024-11-06_08:40:00_hourly``` 

334 

335```$ zfs list -t snapshot -r tank2/boo/bar 

336 

337tank2/boo/bar@prod_us-west-1_2024-11-06_08:30:05_daily 

338 

339tank2/boo/bar@prod_us-west-1_2024-11-06_08:30:05_hourly 

340 

341tank2/boo/bar/baz@prod_us-west-1_2024-11-06_08:40:00_daily 

342 

343tank2/boo/bar/baz@prod_us-west-1_2024-11-06_08:40:00_hourly``` 

344 

345* Example that makes destination identical to source even if the two have drastically diverged: 

346 

347```$ {prog_name} tank1/foo/bar tank2/boo/bar --recursive --force --delete-dst-datasets --delete-dst-snapshots``` 

348 

349* Replicate all daily snapshots created during the last 7 days, and at the same time ensure that the latest 7 daily 

350snapshots (per dataset) are replicated regardless of creation time: 

351 

352```$ {prog_name} tank1/foo/bar tank2/boo/bar --recursive --include-snapshot-regex '.*_daily' 

353--include-snapshot-times-and-ranks '7 days ago..anytime' 'latest 7'``` 

354 

355Note: The example above compares the specified times against the standard ZFS 'creation' time property of the snapshots 

356(which is a UTC Unix time in integer seconds), rather than against a timestamp that may be part of the snapshot name. 

357 

358* Delete all daily snapshots older than 7 days, but ensure that the latest 7 daily snapshots (per dataset) are retained 

359regardless of creation time: 

360 

361```$ {prog_name} {dummy_dataset} tank2/boo/bar --dryrun --recursive --skip-replication --delete-dst-snapshots 

362--include-snapshot-regex '.*_daily' --include-snapshot-times-and-ranks notime 'all except latest 7' 

363--include-snapshot-times-and-ranks 'anytime..7 days ago'``` 

364 

365Note: This also prints how many GB of disk space in total would be freed if the command were to be run for real without 

366the --dryrun flag. 

367 

368* Delete all daily snapshots older than 7 days, but ensure that the latest 7 daily snapshots (per dataset) are retained 

369regardless of creation time. Additionally, only delete a snapshot if no corresponding snapshot or bookmark exists in 

370the source dataset (same as above except replace the 'dummy' source with 'tank1/foo/bar'): 

371 

372```$ {prog_name} tank1/foo/bar tank2/boo/bar --dryrun --recursive --skip-replication --delete-dst-snapshots 

373--include-snapshot-regex '.*_daily' --include-snapshot-times-and-ranks notime 'all except latest 7' 

374--include-snapshot-times-and-ranks '7 days ago..anytime'``` 

375 

376* Delete all daily snapshots older than 7 days, but ensure that the latest 7 daily snapshots (per dataset) are retained 

377regardless of creation time. Additionally, only delete a snapshot if no corresponding snapshot exists in the source 

378dataset (same as above except append 'no-crosscheck'): 

379 

380```$ {prog_name} tank1/foo/bar tank2/boo/bar --dryrun --recursive --skip-replication --delete-dst-snapshots 

381--include-snapshot-regex '.*_daily' --include-snapshot-times-and-ranks notime 'all except latest 7' 

382--include-snapshot-times-and-ranks 'anytime..7 days ago' --delete-dst-snapshots-no-crosscheck``` 

383 

384* Delete all daily bookmarks older than 90 days, but retain the latest 200 daily bookmarks (per dataset) regardless 

385of creation time: 

386 

387```$ {prog_name} {dummy_dataset} tank1/foo/bar --dryrun --recursive --skip-replication --delete-dst-snapshots=bookmarks 

388--include-snapshot-regex '.*_daily' --include-snapshot-times-and-ranks notime 'all except latest 200' 

389--include-snapshot-times-and-ranks 'anytime..90 days ago'``` 

390 

391* Delete all tmp datasets within tank2/boo/bar: 

392 

393```$ {prog_name} {dummy_dataset} tank2/boo/bar --dryrun --recursive --skip-replication --delete-dst-datasets 

394--include-dataset-regex '(.*/)?tmp.*' --exclude-dataset-regex '!.*'``` 

395 

396* Retain all secondly snapshots that were created less than 40 seconds ago, and ensure that the latest 40 

397secondly snapshots (per dataset) are retained regardless of creation time. Same for 40 minutely snapshots, 36 hourly 

398snapshots, 31 daily snapshots, 12 weekly snapshots, 18 monthly snapshots, and 5 yearly snapshots: 

399 

400```$ {prog_name} {dummy_dataset} tank2/boo/bar --dryrun --recursive --skip-replication --delete-dst-snapshots 

401--delete-dst-snapshots-except 

402--include-snapshot-regex '.*_secondly' --include-snapshot-times-and-ranks '40 seconds ago..anytime' 'latest 40' 

403--new-snapshot-filter-group 

404--include-snapshot-regex '.*_minutely' --include-snapshot-times-and-ranks '40 minutes ago..anytime' 'latest 40' 

405--new-snapshot-filter-group 

406--include-snapshot-regex '.*_hourly' --include-snapshot-times-and-ranks '36 hours ago..anytime' 'latest 36' 

407--new-snapshot-filter-group 

408--include-snapshot-regex '.*_daily' --include-snapshot-times-and-ranks '31 days ago..anytime' 'latest 31' 

409--new-snapshot-filter-group 

410--include-snapshot-regex '.*_weekly' --include-snapshot-times-and-ranks '12 weeks ago..anytime' 'latest 12' 

411--new-snapshot-filter-group 

412--include-snapshot-regex '.*_monthly' --include-snapshot-times-and-ranks '18 months ago..anytime' 'latest 18' 

413--new-snapshot-filter-group 

414--include-snapshot-regex '.*_yearly' --include-snapshot-times-and-ranks '5 years ago..anytime' 'latest 5'``` 

415 

416For convenience, the lengthy command line above can be expressed in a more concise way, like so: 

417 

418```$ {prog_name} {dummy_dataset} tank2/boo/bar --dryrun --recursive --skip-replication --delete-dst-snapshots 

419--delete-dst-snapshots-except-plan "{delete_dst_snapshots_except_plan_example1}"``` 

420 

421* Compare source and destination dataset trees recursively, for example to check if all recently taken snapshots have 

422been successfully replicated by a periodic job. List snapshots only contained in src (tagged with 'src'), 

423only contained in dst (tagged with 'dst'), and contained in both src and dst (tagged with 'all'), restricted to hourly 

424and daily snapshots taken within the last 7 days, excluding the last 4 hours (to allow for some slack/stragglers), 

425excluding temporary datasets: 

426 

427```$ {prog_name} tank1/foo/bar tank2/boo/bar --skip-replication --compare-snapshot-lists=src+dst+all --recursive 

428--include-snapshot-regex '.*_(hourly|daily)' --include-snapshot-times-and-ranks '7 days ago..4 hours ago' 

429--exclude-dataset-regex '(.*/)?tmp.*'``` 

430 

431If the resulting TSV output file contains zero lines starting with the prefix 'src' and zero lines starting with the 

432prefix 'dst' then no source snapshots are missing on the destination, and no destination snapshots are missing 

433on the source, indicating that the periodic replication and pruning jobs perform as expected. The TSV output is sorted 

434by dataset, and by ZFS creation time within each dataset - the first and last line prefixed with 'all' contains the 

435metadata of the oldest and latest common snapshot, respectively. The --compare-snapshot-lists option also directly 

436logs various summary stats, such as the metadata of the latest common snapshot, latest snapshots and oldest snapshots, 

437as well as the time diff between the latest common snapshot and latest snapshot only in src (and only in dst), as well 

438as how many src snapshots and how many GB of data are missing on dst, etc. 

439 

440* Example with further options: 

441 

442```$ {prog_name} tank1/foo/bar root@host2.example.com:tank2/boo/bar --recursive 

443--exclude-snapshot-regex '.*_(secondly|minutely)' --exclude-snapshot-regex 'test_.*' 

444--include-snapshot-times-and-ranks '7 days ago..anytime' 'latest 7' --exclude-dataset /tank1/foo/bar/temporary 

445--exclude-dataset /tank1/foo/bar/baz/trash --exclude-dataset-regex '(.*/)?private' 

446--exclude-dataset-regex '(.*/)?[Tt][Ee]?[Mm][Pp][-_]?[0-9]*' --ssh-dst-private-key /root/.ssh/id_rsa``` 

447""") 

448 

449 parser.add_argument( 

450 "root_dataset_pairs", nargs="+", action=DatasetPairsAction, metavar="SRC_DATASET DST_DATASET", 

451 help="SRC_DATASET: " 

452 "Source ZFS dataset (and its descendants) that will be replicated. Can be a ZFS filesystem or ZFS volume. " 

453 "Format is [[user@]host:]dataset. The host name can also be an IPv4 address (or an IPv6 address where " 

454 "each ':' colon character must be replaced with a '|' pipe character for disambiguation). If the " 

455 "host name is '-', the dataset will be on the local host, and the corresponding SSH leg will be omitted. " 

456 "The same is true if the host is omitted and the dataset does not contain a ':' colon at the same time. " 

457 "Local dataset examples: `tank1/foo/bar`, `tank1`, `-:tank1/foo/bar:baz:boo` " 

458 "Remote dataset examples: `host:tank1/foo/bar`, `host.example.com:tank1/foo/bar`, " 

459 "`root@host:tank`, `root@host.example.com:tank1/foo/bar`, `user@127.0.0.1:tank1/foo/bar:baz:boo`, " 

460 "`user@||1:tank1/foo/bar:baz:boo`. " 

461 "The first component of the ZFS dataset name is the ZFS pool name, here `tank1`. " 

462 "If the option starts with a `+` prefix then dataset names are read from the UTF-8 text file given " 

463 "after the `+` prefix, with each line in the file containing a SRC_DATASET and a DST_DATASET, " 

464 "separated by a tab character. Example: `+root_dataset_names.txt`, `+/path/to/root_dataset_names.txt`\n\n" 

465 "DST_DATASET: " 

466 "Destination ZFS dataset for replication and deletion. Has same naming format as SRC_DATASET. During " 

467 "replication, destination datasets that do not yet exist are created as necessary, along with their " 

468 "parent and ancestors.\n\n" 

469 f"*Performance Note:* {prog_name} automatically replicates multiple datasets in parallel. It replicates " 

470 "snapshots in parallel across datasets and serially within a dataset. All child datasets of a dataset " 

471 "may be processed in parallel. For consistency, processing of a dataset only starts after processing of " 

472 "all its ancestor datasets has completed. Further, when a thread is ready to start processing another " 

473 "dataset, it chooses the next dataset wrt. lexicographical sort order from the datasets that are " 

474 "currently available for start of processing. Initially, only the roots of the selected dataset subtrees " 

475 "are available for start of processing. The degree of parallelism is configurable with the --threads " 

476 "option (see below).\n\n") 

477 parser.add_argument( 

478 "--recursive", "-r", action="store_true", 

479 help="During snapshot creation, replication, deletion and comparison, also consider descendant datasets, i.e. " 

480 "datasets within the dataset tree, including children, and children of children, etc.\n\n") 

481 parser.add_argument( 

482 "--include-dataset", action=FileOrLiteralAction, nargs="+", default=[], metavar="DATASET", 

483 help="During snapshot creation, replication, deletion and comparison, select any ZFS dataset (and its descendants) " 

484 "that is contained within SRC_DATASET (DST_DATASET in case of deletion) if its dataset name is one of the " 

485 "given include dataset names but none of the exclude dataset names. If a dataset is excluded its descendants " 

486 "are automatically excluded too, and this decision is never reconsidered even for the descendants because " 

487 "exclude takes precedence over include.\n\n" 

488 "A dataset name is absolute if the specified dataset is prefixed by `/`, e.g. `/tank/baz/tmp`. " 

489 "Otherwise the dataset name is relative wrt. source and destination, e.g. `baz/tmp` if the source " 

490 "is `tank`.\n\n" 

491 "This option is automatically translated to an --include-dataset-regex (see below) and can be " 

492 "specified multiple times.\n\n" 

493 "If the option starts with a `+` prefix then dataset names are read from the newline-separated " 

494 "UTF-8 text file given after the `+` prefix, one dataset per line inside of the text file.\n\n" 

495 "Examples: `/tank/baz/tmp` (absolute), `baz/tmp` (relative), " 

496 "`+dataset_names.txt`, `+/path/to/dataset_names.txt`\n\n") 

497 parser.add_argument( 

498 "--exclude-dataset", action=FileOrLiteralAction, nargs="+", default=[], metavar="DATASET", 

499 help="Same syntax as --include-dataset (see above) except that the option is automatically translated to an " 

500 "--exclude-dataset-regex (see below).\n\n") 

501 parser.add_argument( 

502 "--include-dataset-regex", action=FileOrLiteralAction, nargs="+", default=[], metavar="REGEX", 

503 help="During snapshot creation, replication (and deletion) and comparison, select any ZFS dataset (and its " 

504 "descendants) that is contained within SRC_DATASET (DST_DATASET in case of deletion) if its relative dataset " 

505 "path (e.g. `baz/tmp`) wrt. SRC_DATASET (DST_DATASET in case of deletion) matches at least one of the given " 

506 "include regular expressions but none of the exclude regular expressions. " 

507 "If a dataset is excluded its descendants are automatically excluded too, and this decision is never " 

508 "reconsidered even for the descendants because exclude takes precedence over include.\n\n" 

509 "This option can be specified multiple times. " 

510 "A leading `!` character indicates logical negation, i.e. the regex matches if the regex with the " 

511 "leading `!` character removed does not match.\n\n" 

512 "If the option starts with a `+` prefix then regex names are read from the newline-separated " 

513 "UTF-8 text file given after the `+` prefix, one regex per line inside of the text file.\n\n" 

514 "Default: `.*` (include all datasets).\n\n" 

515 "Examples: `baz/tmp`, `(.*/)?doc[^/]*/(private|confidential).*`, `!public`, " 

516 "`+dataset_regexes.txt`, `+/path/to/dataset_regexes.txt`\n\n") 

517 parser.add_argument( 

518 "--exclude-dataset-regex", action=FileOrLiteralAction, nargs="+", default=[], metavar="REGEX", 

519 help="Same syntax as --include-dataset-regex (see above) except that the default is " 

520 f"`{exclude_dataset_regexes_default}` (exclude tmp datasets). Example: `!.*` (exclude no dataset)\n\n") 

521 parser.add_argument( 

522 "--exclude-dataset-property", default=None, action=NonEmptyStringAction, metavar="STRING", 

523 help="The name of a ZFS dataset user property (optional). If this option is specified, the effective value " 

524 "(potentially inherited) of that user property is read via 'zfs list' for each selected source dataset " 

525 "to determine whether the dataset will be included or excluded, as follows:\n\n" 

526 "a) Value is 'true' or '-' or empty string or the property is missing: Include the dataset.\n\n" 

527 "b) Value is 'false': Exclude the dataset and its descendants.\n\n" 

528 "c) Value is a comma-separated list of host names (no spaces, for example: " 

529 "'store001,store002'): Include the dataset if the host name of " 

530 f"the host executing {prog_name} is contained in the list, otherwise exclude the dataset and its " 

531 "descendants.\n\n" 

532 "If a dataset is excluded its descendants are automatically excluded too, and the property values of the " 

533 "descendants are ignored because exclude takes precedence over include.\n\n" 

534 "Examples: 'syncoid:sync', 'com.example.eng.project.x:backup'\n\n" 

535 "*Note:* The use of --exclude-dataset-property is discouraged for most use cases. It is more flexible, " 

536 "more powerful, *and* more efficient to instead use a combination of --include/exclude-dataset-regex " 

537 "and/or --include/exclude-dataset to achieve the same or better outcome.\n\n") 

538 parser.add_argument( 

539 "--include-snapshot-regex", action=FileOrLiteralAction, nargs="+", default=[], metavar="REGEX", 

540 help="During replication, deletion and comparison, select any source ZFS snapshot that has a name (i.e. the part " 

541 "after the '@') that matches at least one of the given include regular expressions but none of the " 

542 "exclude regular expressions. If a snapshot is excluded this decision is never reconsidered because " 

543 "exclude takes precedence over include.\n\n" 

544 "This option can be specified multiple times. " 

545 "A leading `!` character indicates logical negation, i.e. the regex matches if the regex with the " 

546 "leading `!` character removed does not match.\n\n" 

547 "Default: `.*` (include all snapshots). " 

548 "Examples: `test_.*`, `!prod_.*`, `.*_(hourly|frequent)`, `!.*_(weekly|daily)`\n\n" 

549 "*Note:* All --include/exclude-snapshot-* CLI option groups are combined into a mini filter pipeline. " 

550 "A filter pipeline is executed in the order given on the command line, left to right. For example if " 

551 "--include-snapshot-times-and-ranks (see below) is specified on the command line before " 

552 "--include/exclude-snapshot-regex, then --include-snapshot-times-and-ranks will be applied before " 

553 "--include/exclude-snapshot-regex. The pipeline results would not always be the same if the order were " 

554 "reversed. Order matters.\n\n" 

555 "*Note:* During replication, bookmarks are always retained aka selected in order to help find common " 

556 "snapshots between source and destination.\n\n") 

557 parser.add_argument( 

558 "--exclude-snapshot-regex", action=FileOrLiteralAction, nargs="+", default=[], metavar="REGEX", 

559 help="Same syntax as --include-snapshot-regex (see above) except that the default is to exclude no " 

560 "snapshots.\n\n") 

561 parser.add_argument( 

562 "--include-snapshot-times-and-ranks", action=TimeRangeAndRankRangeAction, nargs="+", default=[], 

563 metavar=("TIMERANGE", "RANKRANGE"), 

564 help="This option takes as input parameters a time range filter and an optional rank range filter. It " 

565 "separately computes the results for each filter and selects the UNION of both results. " 

566 "To instead use a pure rank range filter (no UNION), or a pure time range filter (no UNION), simply " 

567 "use 'notime' aka '0..0' to indicate an empty time range, or omit the rank range, respectively. " 

568 "This option can be specified multiple times.\n\n" 

569 "<b>*Replication Example (UNION):* </b>\n\n" 

570 "Specify to replicate all daily snapshots created during the last 7 days, " 

571 "and at the same time ensure that the latest 7 daily snapshots (per dataset) are replicated regardless " 

572 "of creation time, like so: " 

573 "`--include-snapshot-regex '.*_daily' --include-snapshot-times-and-ranks '7 days ago..anytime' 'latest 7'`\n\n" 

574 "<b>*Deletion Example (no UNION):* </b>\n\n" 

575 "Specify to delete all daily snapshots older than 7 days, but ensure that the " 

576 "latest 7 daily snapshots (per dataset) are retained regardless of creation time, like so: " 

577 "`--include-snapshot-regex '.*_daily' --include-snapshot-times-and-ranks notime 'all except latest 7' " 

578 "--include-snapshot-times-and-ranks 'anytime..7 days ago'`" 

579 "\n\n" 

580 "This helps to safely cope with irregular scenarios where no snapshots were created or received within " 

581 "the last 7 days, or where more than 7 daily snapshots were created within the last 7 days. It can also " 

582 "help to avoid accidental pruning of the last snapshot that source and destination have in common.\n\n" 

583 "" 

584 "<b>*TIMERANGE:* </b>\n\n" 

585 "The ZFS 'creation' time of a snapshot (and bookmark) must fall into this time range in order for the " 

586 "snapshot to be included. The time range consists of a 'start' time, followed by a '..' separator, " 

587 "followed by an 'end' time. For example '2024-01-01..2024-04-01', or 'anytime..anytime' aka `*..*` aka all " 

588 "times, or 'notime' aka '0..0' aka empty time range. Only snapshots (and bookmarks) in the half-open time " 

589 "range [start, end) are included; other snapshots (and bookmarks) are excluded. If a snapshot is excluded " 

590 "this decision is never reconsidered because exclude takes precedence over include. Each of the two specified " 

591 "times can take any of the following forms:\n\n" 

592 "* a) `anytime` aka `*` wildcard; represents negative or positive infinity.\n\n" 

593 "* b) a non-negative integer representing a UTC Unix time in seconds. Example: 1728109805\n\n" 

594 "* c) an ISO 8601 datetime string with or without timezone. Examples: '2024-10-05', " 

595 "'2024-10-05T14:48:55', '2024-10-05T14:48:55+02', '2024-10-05T14:48:55-04:30'. If the datetime string " 

596 "does not contain time zone info then it is assumed to be in the local time zone. Timezone string support " 

597 "requires Python >= 3.11.\n\n" 

598 "* d) a duration that indicates how long ago from the current time, using the following syntax: " 

599 "a non-negative integer, followed by an optional space, followed by a duration unit that is " 

600 "*one* of 'seconds', 'secs', 'minutes', 'mins', 'hours', 'days', 'weeks', 'months', 'years', " 

601 "followed by an optional space, followed by the word 'ago'. " 

602 "Examples: '0secs ago', '40 mins ago', '36hours ago', '90days ago', '12weeksago'.\n\n" 

603 "* Note: This option compares the specified time against the standard ZFS 'creation' time property of the " 

604 "snapshot (which is a UTC Unix time in integer seconds), rather than against a timestamp that may be " 

605 "part of the snapshot name. You can list the ZFS creation time of snapshots and bookmarks as follows: " 

606 "`zfs list -t snapshot,bookmark -o name,creation -s creation -d 1 $SRC_DATASET` (optionally add " 

607 "the -p flag to display UTC Unix time in integer seconds).\n\n" 

608 "*Note:* During replication, bookmarks are always retained aka selected in order to help find common " 

609 "snapshots between source and destination.\n\n" 

610 "" 

611 "<b>*RANKRANGE:* </b>\n\n" 

612 "Specifies to include the N (or N%%) oldest snapshots or latest snapshots, and exclude all other " 

613 "snapshots (default: include no snapshots). Snapshots are sorted by creation time (actually, by the " 

614 "'createtxg' ZFS property, which serves the same purpose but is more precise). The rank position of a " 

615 "snapshot is the zero-based integer position of the snapshot within that sorted list. A rank consists of the " 

616 "optional words 'all except' (followed by an optional space), followed by the word 'oldest' or 'latest', " 

617 "followed by a non-negative integer, followed by an optional '%%' percent sign. A rank range consists of a " 

618 "lower rank, followed by a '..' separator, followed by a higher rank. " 

619 "If the optional lower rank is missing it is assumed to be 0. Examples:\n\n" 

620 "* 'oldest 10%%' aka 'oldest 0..oldest 10%%' (include the oldest 10%% of all snapshots)\n\n" 

621 "* 'latest 10%%' aka 'latest 0..latest 10%%' (include the latest 10%% of all snapshots)\n\n" 

622 "* 'all except latest 10%%' aka 'oldest 90%%' aka 'oldest 0..oldest 90%%' (include all snapshots except the " 

623 "latest 10%% of all snapshots)\n\n" 

624 "* 'oldest 90' aka 'oldest 0..oldest 90' (include the oldest 90 snapshots)\n\n" 

625 "* 'latest 90' aka 'latest 0..latest 90' (include the latest 90 snapshots)\n\n" 

626 "* 'all except oldest 90' aka 'oldest 90..oldest 100%%' (include all snapshots except the oldest 90 snapshots)" 

627 "\n\n" 

628 "* 'all except latest 90' aka 'latest 90..latest 100%%' (include all snapshots except the latest 90 snapshots)" 

629 "\n\n" 

630 "* 'latest 1' aka 'latest 0..latest 1' (include the latest snapshot)\n\n" 

631 "* 'all except latest 1' aka 'latest 1..latest 100%%' (include all snapshots except the latest snapshot)\n\n" 

632 "* 'oldest 2' aka 'oldest 0..oldest 2' (include the oldest 2 snapshots)\n\n" 

633 "* 'all except oldest 2' aka 'oldest 2..oldest 100%%' (include all snapshots except the oldest 2 snapshots)\n\n" 

634 "* 'oldest 100%%' aka 'oldest 0..oldest 100%%' (include all snapshots)\n\n" 

635 "* 'oldest 0%%' aka 'oldest 0..oldest 0%%' (include no snapshots)\n\n" 

636 "* 'oldest 0' aka 'oldest 0..oldest 0' (include no snapshots)\n\n" 

637 "*Note:* If multiple RANKRANGEs are specified within a single --include-snapshot-times-and-ranks option, each " 

638 "subsequent rank range operates on the output of the preceding rank rage.\n\n" 

639 "*Note:* Percentage calculations are not based on the number of snapshots " 

640 "contained in the dataset on disk, but rather based on the number of snapshots arriving at the filter. " 

641 "For example, if only two daily snapshots arrive at the filter because a prior filter excludes hourly " 

642 "snapshots, then 'latest 10' will only include these two daily snapshots, and 'latest 50%%' will only " 

643 "include one of these two daily snapshots.\n\n" 

644 "*Note:* During replication, bookmarks are always retained aka selected in order to help find common " 

645 "snapshots between source and destination. Bookmarks do not count towards N or N%% wrt. rank.\n\n" 

646 "*Note:* If a snapshot is excluded this decision is never reconsidered because exclude takes precedence " 

647 "over include.\n\n") 

648 

649 src_snapshot_plan_example = { 

650 "prod": { 

651 "onsite": {"secondly": 40, "minutely": 40, "hourly": 36, "daily": 31, "weekly": 12, "monthly": 18, "yearly": 5}, 

652 "us-west-1": {"secondly": 0, "minutely": 0, "hourly": 36, "daily": 31, "weekly": 12, "monthly": 18, "yearly": 5}, 

653 "eu-west-1": {"secondly": 0, "minutely": 0, "hourly": 36, "daily": 31, "weekly": 12, "monthly": 18, "yearly": 5}, 

654 }, 

655 "test": { 

656 "offsite": {"12hourly": 42, "weekly": 12}, 

657 "onsite": {"100millisecondly": 42}, 

658 }, 

659 } 

660 parser.add_argument( 

661 "--include-snapshot-plan", action=IncludeSnapshotPlanAction, default=None, metavar="DICT_STRING", 

662 help="Replication periods to be used if replicating snapshots within the selected destination datasets. " 

663 "Has the same format as --create-src-snapshots-plan and --delete-dst-snapshots-except-plan (see below). " 

664 "Snapshots that do not match a period will not be replicated. To avoid unexpected surprises, make sure to " 

665 "carefully specify ALL snapshot names and periods that shall be replicated, in combination with --dryrun.\n\n" 

666 f"Example: `{format_dict(src_snapshot_plan_example)}`. This example will, for the organization 'prod' and the " 

667 "intended logical target 'onsite', replicate secondly snapshots that were created less than 40 seconds ago, " 

668 "yet replicate the latest 40 secondly snapshots regardless of creation time. Analog for the latest 40 minutely " 

669 "snapshots, latest 36 hourly snapshots, etc. " 

670 "Note: A zero within a period (e.g. 'hourly': 0) indicates that no snapshots shall be replicated for the given " 

671 "period.\n\n" 

672 "Note: --include-snapshot-plan is a convenience option that auto-generates a series of the following other " 

673 "options: --new-snapshot-filter-group, --include-snapshot-regex, --include-snapshot-times-and-ranks\n\n") 

674 parser.add_argument( 

675 "--new-snapshot-filter-group", action=NewSnapshotFilterGroupAction, nargs=0, 

676 help="Starts a new snapshot filter group containing separate --{include|exclude}-snapshot-* filter options. The " 

677 "program separately computes the results for each filter group and selects the UNION of all results. " 

678 "This option can be specified multiple times and serves as a separator between groups. Example:\n\n" 

679 "Delete all minutely snapshots older than 40 minutes, but ensure that the latest 40 minutely snapshots (per " 

680 "dataset) are retained regardless of creation time. Additionally, delete all hourly snapshots older than 36 " 

681 "hours, but ensure that the latest 36 hourly snapshots (per dataset) are retained regardless of creation time. " 

682 "Additionally, delete all daily snapshots older than 31 days, but ensure that the latest 31 daily snapshots " 

683 "(per dataset) are retained regardless of creation time: " 

684 f"`{prog_name} {dummy_dataset} tank2/boo/bar --dryrun --recursive --skip-replication --delete-dst-snapshots " 

685 "--include-snapshot-regex '.*_minutely' --include-snapshot-times-and-ranks notime 'all except latest 40' " 

686 "--include-snapshot-times-and-ranks 'anytime..40 minutes ago' " 

687 "--new-snapshot-filter-group " 

688 "--include-snapshot-regex '.*_hourly' --include-snapshot-times-and-ranks notime 'all except latest 36' " 

689 "--include-snapshot-times-and-ranks 'anytime..36 hours ago' " 

690 "--new-snapshot-filter-group " 

691 "--include-snapshot-regex '.*_daily' --include-snapshot-times-and-ranks notime 'all except latest 31' " 

692 "--include-snapshot-times-and-ranks 'anytime..31 days ago'`\n\n") 

693 parser.add_argument( 

694 "--create-src-snapshots", action="store_true", 

695 help="Do nothing if the --create-src-snapshots flag is missing. Otherwise, before the replication step (see below), " 

696 "atomically create new snapshots of the source datasets selected via --{include|exclude}-dataset* policy. " 

697 "The names of the snapshots can be configured via --create-src-snapshots-* suboptions (see below). " 

698 "To create snapshots only, without any other processing such as replication, etc, consider using this flag " 

699 "together with the --skip-replication flag.\n\n" 

700 "A periodic snapshot is created if it is due per the schedule indicated by --create-src-snapshots-plan " 

701 "(for example '_daily' or '_hourly' or _'10minutely' or '_2secondly' or '_100millisecondly'), or if the " 

702 "--create-src-snapshots-even-if-not-due flag is specified, or if the most recent scheduled snapshot " 

703 f"is somehow missing. In the latter case {prog_name} immediately creates a snapshot (tagged with the current " 

704 "time, not backdated to the missed time), and then resumes the original schedule.\n\n" 

705 "If the snapshot suffix is '_adhoc' or not a known period then a snapshot is considered " 

706 "non-periodic and is thus created immediately regardless of the creation time of any existing snapshot.\n\n" 

707 "The implementation attempts to fit as many datasets as possible into a single (atomic) 'zfs snapshot' command " 

708 "line, using lexicographical sort order, and using 'zfs snapshot -r' to the extent that this is compatible " 

709 "with the actual results of the schedule and the actual results of the --{include|exclude}-dataset* pruning " 

710 "policy. The snapshots of all datasets that fit " 

711 "within the same single 'zfs snapshot' CLI invocation will be taken within the same ZFS transaction group, and " 

712 "correspondingly have identical 'createtxg' ZFS property (but not necessarily identical 'creation' ZFS time " 

713 "property as ZFS actually provides no such guarantee), and thus be consistent. Dataset names that can't fit " 

714 "into a single command line are spread over multiple command line invocations, respecting the limits that the " 

715 "operating system places on the maximum length of a single command line, per `getconf ARG_MAX`.\n\n" 

716 f"Note: All {prog_name} functions including snapshot creation, replication, deletion, monitoring, comparison, " 

717 "etc. happily work with any snapshots in any format, even created or managed by third party ZFS snapshot " 

718 "management tools, including manual zfs snapshot/destroy.\n\n") 

719 parser.add_argument( 

720 "--create-src-snapshots-plan", default=None, type=str, metavar="DICT_STRING", 

721 help="Creation periods that specify a schedule for when new snapshots shall be created on src within the selected " 

722 "datasets. Has the same format as --delete-dst-snapshots-except-plan.\n\n" 

723 f"Example: `{format_dict(src_snapshot_plan_example)}`. This example will, for the organization 'prod' and " 

724 "the intended logical target 'onsite', create 'secondly' snapshots every second, 'minutely' snapshots every " 

725 "minute, hourly snapshots every hour, and so on. " 

726 "It will also create snapshots for the targets 'us-west-1' and 'eu-west-1' within the 'prod' organization. " 

727 "In addition, it will create snapshots every 12 hours and every week for the 'test' organization, " 

728 "and name them as being intended for the 'offsite' replication target. Analog for snapshots that are taken " 

729 "every 100 milliseconds within the 'test' organization.\n\n" 

730 "The example creates ZFS snapshots with names like " 

731 "`prod_onsite_<timestamp>_secondly`, `prod_onsite_<timestamp>_minutely`, " 

732 "`prod_us-west-1_<timestamp>_hourly`, `prod_us-west-1_<timestamp>_daily`, " 

733 "`prod_eu-west-1_<timestamp>_hourly`, `prod_eu-west-1_<timestamp>_daily`, " 

734 "`test_offsite_<timestamp>_12hourly`, `test_offsite_<timestamp>_weekly`, and so on.\n\n" 

735 "Note: A period name that is missing indicates that no snapshots shall be created for the given period.\n\n" 

736 "The period name can contain an optional positive integer immediately preceding the time period unit, for " 

737 "example `_2secondly` or `_10minutely` or `_100millisecondly` to indicate that snapshots are taken every 2 " 

738 "seconds, or every 10 minutes, or every 100 milliseconds, respectively.\n\n") 

739 

740 def argparser_escape(text: str) -> str: 

741 return text.replace('%', '%%') 

742 

743 parser.add_argument( 

744 "--create-src-snapshots-timeformat", default="%Y-%m-%d_%H:%M:%S", metavar="STRFTIME_SPEC", 

745 help="Default is `%(default)s`. For the strftime format, see " 

746 "https://docs.python.org/3.11/library/datetime.html#strftime-strptime-behavior. " 

747 f"Examples: `{argparser_escape('%Y-%m-%d_%H:%M:%S.%f')}` (adds microsecond resolution), " 

748 f"`{argparser_escape('%Y-%m-%d_%H:%M:%S%z')}` (adds timezone offset), " 

749 f"`{argparser_escape('%Y-%m-%dT%H-%M-%S')}` (no colons).\n\n" 

750 "The name of the snapshot created on the src is `$org_$target_strftime(--create-src-snapshots-time*)_$period`. " 

751 "Example: `tank/foo@prod_us-west-1_2024-09-03_12:26:15_daily`\n\n") 

752 parser.add_argument( 

753 "--create-src-snapshots-timezone", default="", type=str, metavar="TZ_SPEC", 

754 help=f"Default is the local timezone of the system running {prog_name}. When creating a new snapshot on the source, " 

755 "fetch the current time in the specified timezone, and feed that time, and the value of " 

756 "--create-src-snapshots-timeformat, into the standard strftime() function to generate the timestamp portion " 

757 "of the snapshot name. The TZ_SPEC input parameter is of the form 'UTC' or '+HHMM' or '-HHMM' for fixed UTC " 

758 "offsets, or an IANA TZ identifier for auto-adjustment to daylight savings time, or the empty string to use " 

759 "the local timezone, for example '', 'UTC', '+0000', '+0530', '-0400', 'America/Los_Angeles', 'Europe/Vienna'. " 

760 "For a list of valid IANA TZ identifiers see https://en.wikipedia.org/wiki/List_of_tz_database_time_zones#List" 

761 "\n\nTo change the timezone not only for snapshot name creation, but in all respects for the entire program, " 

762 "use the standard 'TZ' Unix environment variable, like so: `export TZ=UTC`.\n\n") 

763 parser.add_argument( 

764 "--create-src-snapshots-even-if-not-due", action="store_true", 

765 help="Take snapshots immediately regardless of the creation time of any existing snapshot, even if snapshots " 

766 "are periodic and not actually due per the schedule.\n\n") 

767 parser.add_argument( 

768 "--create-src-snapshots-enable-snapshots-changed-cache", action="store_true", 

769 help=argparse.SUPPRESS) # deprecated; was replaced by --cache-snapshots 

770 parser.add_argument( 

771 "--zfs-send-program-opts", type=str, default="--props --raw --compressed", metavar="STRING", 

772 help="Parameters to fine-tune 'zfs send' behaviour (optional); will be passed into 'zfs send' CLI. " 

773 "The value is split on runs of one or more whitespace characters. " 

774 "Default is '%(default)s'. To run `zfs send` without options, specify the empty " 

775 "string: `--zfs-send-program-opts=''`. " 

776 "See https://openzfs.github.io/openzfs-docs/man/master/8/zfs-send.8.html " 

777 "and https://github.com/openzfs/zfs/issues/13024\n\n") 

778 parser.add_argument( 

779 "--zfs-recv-program-opts", type=str, default="-u", metavar="STRING", 

780 help="Parameters to fine-tune 'zfs receive' behaviour (optional); will be passed into 'zfs receive' CLI. " 

781 "The value is split on runs of one or more whitespace characters. " 

782 "Default is '%(default)s'. To run `zfs receive` without options, specify the empty " 

783 "string: `--zfs-recv-program-opts=''`. " 

784 "Example: '-u -o canmount=noauto -o readonly=on -x keylocation -x keyformat -x encryption'. " 

785 "See https://openzfs.github.io/openzfs-docs/man/master/8/zfs-receive.8.html " 

786 "and https://openzfs.github.io/openzfs-docs/man/master/7/zfsprops.7.html\n\n") 

787 parser.add_argument( 

788 "--zfs-recv-program-opt", action="append", default=[], metavar="STRING", 

789 help="Parameter to fine-tune 'zfs receive' behaviour (optional); will be passed into 'zfs receive' CLI. " 

790 "The value can contain spaces and is not split. This option can be specified multiple times. Example: `" 

791 "--zfs-recv-program-opt=-o " 

792 "--zfs-recv-program-opt='org.zfsbootmenu:commandline=ro debug zswap.enabled=1'`\n\n") 

793 parser.add_argument( 

794 "--force-rollback-to-latest-snapshot", action="store_true", 

795 help="Before replication, rollback the destination dataset to its most recent destination snapshot (if there " 

796 "is one), via 'zfs rollback', just in case the destination dataset was modified since its most recent " 

797 "snapshot. This is much less invasive than the other --force* options (see below).\n\n") 

798 parser.add_argument( 

799 "--force-rollback-to-latest-common-snapshot", action="store_true", 

800 help="Before replication, delete destination ZFS snapshots that are more recent than the most recent common " 

801 "snapshot selected on the source ('conflicting snapshots'), via 'zfs rollback'. Do no rollback if no common " 

802 "snapshot is selected.\n\n") 

803 parser.add_argument( 

804 "--force", action="store_true", 

805 help="Same as --force-rollback-to-latest-common-snapshot (see above), except that additionally, if no common " 

806 "snapshot is selected, then delete all destination snapshots before starting replication, and proceed " 

807 "without aborting. Without the --force* flags, the destination dataset is treated as append-only, hence " 

808 "no destination snapshot that already exists is deleted, and instead the operation is aborted with an " 

809 "error when encountering a conflicting snapshot.\n\n" 

810 "Analogy: --force-rollback-to-latest-snapshot is a tiny hammer, whereas " 

811 "--force-rollback-to-latest-common-snapshot is a medium sized hammer, --force is a large hammer, and " 

812 "--force-destroy-dependents is a very large hammer. " 

813 "Consider using the smallest hammer that can fix the problem. No hammer is ever used by default.\n\n") 

814 parser.add_argument( 

815 "--force-destroy-dependents", action="store_true", 

816 help="On destination, --force and --force-rollback-to-latest-common-snapshot and --delete-* will add the " 

817 "'-R' flag to their use of 'zfs rollback' and 'zfs destroy', causing them to delete dependents such as " 

818 "clones and bookmarks. This can be very destructive and is rarely advisable.\n\n") 

819 parser.add_argument( 

820 "--force-hard", action="store_true", # deprecated; was renamed to --force-destroy-dependents 

821 help=argparse.SUPPRESS) 

822 parser.add_argument( 

823 "--force-unmount", action="store_true", 

824 help="On destination, --force and --force-rollback-to-latest-common-snapshot will add the '-f' flag to their " 

825 "use of 'zfs rollback' and 'zfs destroy'.\n\n") 

826 parser.add_argument( 

827 "--force-once", "--f1", action="store_true", 

828 help="Use the --force option or --force-rollback-to-latest-common-snapshot option at most once to resolve a " 

829 "conflict, then abort with an error on any subsequent conflict. This helps to interactively resolve " 

830 "conflicts, one conflict at a time.\n\n") 

831 parser.add_argument( 

832 "--skip-parent", action="store_true", 

833 help="During replication and deletion, skip processing of the SRC_DATASET and DST_DATASET and only process " 

834 "their descendant datasets, i.e. children, and children of children, etc (with --recursive). No dataset " 

835 "is processed unless --recursive is also specified. " 

836 f"Analogy: `{prog_name} --recursive --skip-parent src dst` is akin to Unix `cp -r src/* dst/` whereas " 

837 f" `{prog_name} --recursive --skip-parent --skip-replication --delete-dst-datasets dummy dst` is akin to " 

838 "Unix `rm -r dst/*`\n\n") 

839 parser.add_argument( 

840 "--skip-missing-snapshots", choices=["fail", "dataset", "continue"], default="dataset", nargs="?", 

841 help="During replication, handle source datasets that select no snapshots (and no relevant bookmarks) " 

842 "as follows:\n\n" 

843 "a) 'fail': Abort with an error.\n\n" 

844 "b) 'dataset' (default): Skip the source dataset with a warning. Skip descendant datasets if " 

845 "--recursive and destination dataset does not exist. Otherwise skip to the next dataset.\n\n" 

846 "c) 'continue': Skip nothing. If destination snapshots exist, delete them (with --force) or abort " 

847 "with an error (without --force). If there is no such abort, continue processing with the next dataset. " 

848 "Eventually create empty destination dataset and ancestors if they do not yet exist and source dataset " 

849 "has at least one descendant that selects at least one snapshot.\n\n") 

850 parser.add_argument( 

851 "--retries", type=int, min=0, default=2, action=CheckRange, metavar="INT", 

852 help="The maximum number of times a retryable replication or deletion step shall be retried if it fails, for " 

853 "example because of network hiccups (default: %(default)s, min: %(min)s). " 

854 "Also consider this option if a periodic pruning script may simultaneously delete a dataset or " 

855 f"snapshot or bookmark while {prog_name} is running and attempting to access it.\n\n") 

856 parser.add_argument( 

857 "--retry-min-sleep-secs", type=float, min=0, default=0.125, action=CheckRange, metavar="FLOAT", 

858 help="The minimum duration to sleep between retries (default: %(default)s).\n\n") 

859 parser.add_argument( 

860 "--retry-max-sleep-secs", type=float, min=0, default=5 * 60, action=CheckRange, metavar="FLOAT", 

861 help="The maximum duration to sleep between retries initially starts with --retry-min-sleep-secs (see above), " 

862 "and doubles on each retry, up to the final maximum of --retry-max-sleep-secs " 

863 "(default: %(default)s). On each retry a random sleep time in the " 

864 "[--retry-min-sleep-secs, current max] range is picked. The timer resets after each operation.\n\n") 

865 parser.add_argument( 

866 "--retry-max-elapsed-secs", type=float, min=0, default=60 * 60, action=CheckRange, metavar="FLOAT", 

867 help="A single operation (e.g. 'zfs send/receive' of the current dataset, or deletion of a list of snapshots " 

868 "within the current dataset) will not be retried (or not retried anymore) once this much time has elapsed " 

869 "since the initial start of the operation, including retries (default: %(default)s). " 

870 "The timer resets after each operation completes or retries exhaust, such that subsequently failing " 

871 "operations can again be retried.\n\n") 

872 parser.add_argument( 

873 "--skip-on-error", choices=["fail", "tree", "dataset"], default="dataset", 

874 help="During replication and deletion, if an error is not retryable, or --retries has been exhausted, " 

875 "or --skip-missing-snapshots raises an error, proceed as follows:\n\n" 

876 "a) 'fail': Abort the program with an error. This mode is ideal for testing, clear " 

877 "error reporting, and situations where consistency trumps availability.\n\n" 

878 "b) 'tree': Log the error, skip the dataset tree rooted at the dataset for which the error " 

879 "occurred, and continue processing the next (sibling) dataset tree. " 

880 "Example: Assume datasets tank/user1/foo and tank/user2/bar and an error occurs while processing " 

881 "tank/user1. In this case processing skips tank/user1/foo and proceeds with tank/user2.\n\n" 

882 "c) 'dataset' (default): Same as 'tree' except if the destination dataset already exists, skip to " 

883 "the next dataset instead.\n\n" 

884 "Example: Assume datasets tank/user1/foo and tank/user2/bar and an error occurs while " 

885 "processing tank/user1. In this case processing skips tank/user1 and proceeds with tank/user1/foo " 

886 "if the destination already contains tank/user1. Otherwise processing continues with tank/user2. " 

887 "This mode is for production use cases that require timely forward progress even in the presence of " 

888 "partial failures. For example, assume the job is to backup the home directories or virtual machines " 

889 "of thousands of users across an organization. Even if replication of some of the datasets for some " 

890 "users fails due too conflicts, busy datasets, etc, the replication job will continue for the " 

891 "remaining datasets and the remaining users.\n\n") 

892 parser.add_argument( 

893 "--skip-replication", action="store_true", 

894 help="Skip replication step (see above) and proceed to the optional --delete-dst-datasets step " 

895 "immediately (see below).\n\n") 

896 parser.add_argument( 

897 "--delete-dst-datasets", action="store_true", 

898 help="Do nothing if the --delete-dst-datasets option is missing. Otherwise, after successful replication " 

899 "step, if any, delete existing destination datasets that are selected via --{include|exclude}-dataset* " 

900 "policy yet do not exist within SRC_DATASET (which can be an empty dataset, such as the hardcoded virtual " 

901 f"dataset named '{dummy_dataset}'!). Do not recurse without --recursive. With --recursive, never delete " 

902 "non-selected dataset subtrees or their ancestors.\n\n" 

903 "For example, if the destination contains datasets h1,h2,h3,d1 whereas source only contains h3, " 

904 "and the include/exclude policy selects h1,h2,h3,d1, then delete datasets h1,h2,d1 on " 

905 "the destination to make it 'the same'. On the other hand, if the include/exclude policy " 

906 "only selects h1,h2,h3 then only delete datasets h1,h2 on the destination to make it 'the same'.\n\n" 

907 "Example to delete all tmp datasets within tank2/boo/bar: " 

908 f"`{prog_name} {dummy_dataset} tank2/boo/bar --dryrun --skip-replication --recursive " 

909 "--delete-dst-datasets --include-dataset-regex '(.*/)?tmp.*' --exclude-dataset-regex '!.*'`\n\n") 

910 parser.add_argument( 

911 "--delete-dst-snapshots", choices=["snapshots", "bookmarks"], default=None, const="snapshots", nargs="?", 

912 help="Do nothing if the --delete-dst-snapshots option is missing. Otherwise, after successful " 

913 "replication, and successful --delete-dst-datasets step, if any, delete existing destination snapshots " 

914 "whose GUID does not exist within the source dataset (which can be an empty dummy dataset!) if the " 

915 "destination snapshots are selected by the --include/exclude-snapshot-* policy, and the destination " 

916 "dataset is selected via --{include|exclude}-dataset* policy. Does not recurse without --recursive.\n\n" 

917 "For example, if the destination dataset contains snapshots h1,h2,h3,d1 (h=hourly, d=daily) whereas " 

918 "the source dataset only contains snapshot h3, and the include/exclude policy selects " 

919 "h1,h2,h3,d1, then delete snapshots h1,h2,d1 on the destination dataset to make it 'the same'. " 

920 "On the other hand, if the include/exclude policy only selects snapshots h1,h2,h3 then only " 

921 "delete snapshots h1,h2 on the destination dataset to make it 'the same'.\n\n" 

922 "*Note:* To delete snapshots regardless, consider using --delete-dst-snapshots in combination with a " 

923 f"source that is an empty dataset, such as the hardcoded virtual dataset named '{dummy_dataset}', like so:" 

924 f" `{prog_name} {dummy_dataset} tank2/boo/bar --dryrun --skip-replication --delete-dst-snapshots " 

925 "--include-snapshot-regex '.*_daily' --recursive`\n\n" 

926 "*Note:* Use --delete-dst-snapshots=bookmarks to delete bookmarks instead of snapshots, in which " 

927 "case no snapshots are selected and the --{include|exclude}-snapshot-* filter options treat bookmarks as " 

928 "snapshots wrt. selecting.\n\n" 

929 "*Performance Note:* --delete-dst-snapshots operates on multiple datasets in parallel (and serially " 

930 f"within a dataset), using the same dataset order as {prog_name} replication. " 

931 "The degree of parallelism is configurable with the --threads option (see below).\n\n") 

932 parser.add_argument( 

933 "--delete-dst-snapshots-no-crosscheck", action="store_true", 

934 help="This flag indicates that --delete-dst-snapshots=snapshots shall check the source dataset only for " 

935 "a snapshot with the same GUID, and ignore whether a bookmark with the same GUID is present in the " 

936 "source dataset. Similarly, it also indicates that --delete-dst-snapshots=bookmarks shall check the " 

937 "source dataset only for a bookmark with the same GUID, and ignore whether a snapshot with the same GUID " 

938 "is present in the source dataset.\n\n") 

939 parser.add_argument( 

940 "--delete-dst-snapshots-except", action="store_true", 

941 help="This flag indicates that the --include/exclude-snapshot-* options shall have inverted semantics for the " 

942 "--delete-dst-snapshots option, thus deleting all snapshots except for the selected snapshots (within the " 

943 "specified datasets), instead of deleting all selected snapshots (within the specified datasets). In other " 

944 "words, this flag enables to specify which snapshots to retain instead of which snapshots to delete.\n\n" 

945 "*Synchronization vs. Backup*: When a real (non-dummy) source dataset is specified in combination with " 

946 "--delete-dst-snapshots-except, then any destination snapshot retained by the rules above is actually only " 

947 "retained if it also exists in the source dataset - __all other destination snapshots are deleted__. This is " 

948 "great for synchronization use cases but should __NEVER BE USED FOR LONG-TERM ARCHIVAL__. Long-term archival " 

949 "use cases should instead specify the `dummy` source dataset as they require an independent retention policy " 

950 "that is not tied to the current contents of the source dataset.\n\n") 

951 parser.add_argument( 

952 "--delete-dst-snapshots-except-plan", action=DeleteDstSnapshotsExceptPlanAction, default=None, metavar="DICT_STRING", 

953 help="Retention periods to be used if pruning snapshots or bookmarks within the selected destination datasets via " 

954 "--delete-dst-snapshots. Has the same format as --create-src-snapshots-plan. " 

955 "Snapshots (--delete-dst-snapshots=snapshots) or bookmarks (with --delete-dst-snapshots=bookmarks) that " 

956 "do not match a period will be deleted. To avoid unexpected surprises, make sure to carefully specify ALL " 

957 "snapshot names and periods that shall be retained, in combination with --dryrun.\n\n" 

958 f"Example: `{format_dict(src_snapshot_plan_example)}`. This example will, for the organization 'prod' and " 

959 "the intended logical target 'onsite', retain secondly snapshots that were created less than 40 seconds ago, " 

960 "yet retain the latest 40 secondly snapshots regardless of creation time. Analog for the latest 40 minutely " 

961 "snapshots, latest 36 hourly snapshots, etc. " 

962 "It will also retain snapshots for the targets 'us-west-1' and 'eu-west-1' within the 'prod' organization. " 

963 "In addition, within the 'test' organization, it will retain snapshots that are created every 12 hours and " 

964 "every week as specified, and name them as being intended for the 'offsite' replication target. Analog for " 

965 "snapshots that are taken every 100 milliseconds within the 'test' organization. " 

966 "All other snapshots within the selected datasets will be deleted - you've been warned!\n\n" 

967 "The example scans the selected ZFS datasets for snapshots with names like " 

968 "`prod_onsite_<timestamp>_secondly`, `prod_onsite_<timestamp>_minutely`, " 

969 "`prod_us-west-1_<timestamp>_hourly`, `prod_us-west-1_<timestamp>_daily`, " 

970 "`prod_eu-west-1_<timestamp>_hourly`, `prod_eu-west-1_<timestamp>_daily`, " 

971 "`test_offsite_<timestamp>_12hourly`, `test_offsite_<timestamp>_weekly`, and so on, and deletes all snapshots " 

972 "that do not match a retention rule.\n\n" 

973 "Note: A zero within a period (e.g. 'hourly': 0) indicates that no snapshots shall be retained for the given " 

974 "period.\n\n" 

975 "Note: --delete-dst-snapshots-except-plan is a convenience option that auto-generates a series of the " 

976 "following other options: --delete-dst-snapshots-except, " 

977 "--new-snapshot-filter-group, --include-snapshot-regex, --include-snapshot-times-and-ranks\n\n") 

978 parser.add_argument( 

979 "--delete-empty-dst-datasets", choices=["snapshots", "snapshots+bookmarks"], default=None, 

980 const="snapshots+bookmarks", nargs="?", 

981 help="Do nothing if the --delete-empty-dst-datasets option is missing or --recursive is missing. Otherwise, " 

982 "after successful replication " 

983 "step and successful --delete-dst-datasets and successful --delete-dst-snapshots steps, if any, " 

984 "delete any selected destination dataset that has no snapshot and no bookmark if all descendants of " 

985 "that destination dataset are also selected and do not have a snapshot or bookmark either " 

986 "(again, only if the existing destination dataset is selected via --{include|exclude}-dataset* policy). " 

987 "Never delete non-selected dataset subtrees or their ancestors.\n\n" 

988 "For example, if the destination contains datasets h1,d1, and the include/exclude policy " 

989 "selects h1,d1, then check if h1,d1 can be deleted. " 

990 "On the other hand, if the include/exclude policy only selects h1 then only check if h1 can be deleted.\n\n" 

991 "*Note:* Use --delete-empty-dst-datasets=snapshots to delete snapshot-less datasets even if they still " 

992 "contain bookmarks.\n\n") 

993 monitor_snapshot_plan_example = { 

994 "prod": { 

995 "onsite": { 

996 "100millisecondly": {"latest": {"warning": "300 milliseconds", "critical": "2 seconds"}}, 

997 "secondly": {"latest": {"warning": "2 seconds", "critical": "14 seconds"}}, 

998 "minutely": {"latest": {"warning": "30 seconds", "critical": "300 seconds"}}, 

999 "hourly": {"latest": {"warning": "30 minutes", "critical": "300 minutes"}}, 

1000 "daily": {"latest": {"warning": "4 hours", "critical": "8 hours"}}, 

1001 "weekly": {"latest": {"warning": "2 days", "critical": "8 days"}}, 

1002 "monthly": {"latest": {"warning": "2 days", "critical": "8 days"}}, 

1003 "yearly": {"latest": {"warning": "5 days", "critical": "14 days"}}, 

1004 "10minutely": {"latest": {"warning": "0 minutes", "critical": "0 minutes"}}, 

1005 }, 

1006 "": { 

1007 "daily": {"latest": {"warning": "4 hours", "critical": "8 hours"}}, 

1008 }, 

1009 }, 

1010 } 

1011 parser.add_argument( 

1012 "--monitor-snapshots", default="{}", type=str, metavar="DICT_STRING", 

1013 help="Do nothing if the --monitor-snapshots flag is missing. Otherwise, after all other steps, " 

1014 "alert the user if the ZFS 'creation' time property of the latest snapshot for any specified snapshot name " 

1015 "pattern within the selected datasets is too old wrt. the specified age limit. The purpose is to check if " 

1016 "snapshots are successfully taken on schedule, successfully replicated on schedule, and successfully pruned on " 

1017 "schedule. Process exit code is 0, 1, 2 on OK, WARNING, CRITICAL, respectively. " 

1018 f"Example DICT_STRING: `{format_dict(monitor_snapshot_plan_example)}`. " 

1019 "This example alerts the user if the latest src or dst snapshot named `prod_onsite_<timestamp>_hourly` is more " 

1020 "than 30 minutes late (i.e. more than 30+60=90 minutes old) [warning] or more than 300 minutes late (i.e. more " 

1021 "than 300+60=360 minutes old) [critical]. " 

1022 "Analog for the latest snapshot named `prod_<timestamp>_daily`, and so on.\n\n" 

1023 "Note: A duration that is missing or zero (e.g. '0 minutes') indicates that no snapshots shall be checked for " 

1024 "the given snapshot name pattern.\n\n") 

1025 parser.add_argument( 

1026 "--monitor-snapshots-dont-warn", action="store_true", 

1027 help="Log a message for monitoring warnings but nonetheless exit with zero exit code.\n\n") 

1028 parser.add_argument( 

1029 "--monitor-snapshots-dont-crit", action="store_true", 

1030 help="Log a message for monitoring criticals but nonetheless exit with zero exit code.\n\n") 

1031 parser.add_argument(