-
Notifications
You must be signed in to change notification settings - Fork 3
/
ceph.py
1810 lines (1524 loc) · 64.3 KB
/
ceph.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
# Copyright 2014-2015 Canonical Limited.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
# Copyright 2012 Canonical Ltd.
#
# This file is sourced from lp:openstack-charm-helpers
#
# Authors:
# James Page <james.page@ubuntu.com>
# Adam Gandelman <adamg@ubuntu.com>
#
import collections
import errno
import hashlib
import math
import six
import os
import shutil
import json
import time
import uuid
from subprocess import (
check_call,
check_output,
CalledProcessError,
)
from charmhelpers.core.hookenv import (
config,
service_name,
local_unit,
relation_get,
relation_ids,
relation_set,
related_units,
log,
DEBUG,
INFO,
WARNING,
ERROR,
)
from charmhelpers.core.host import (
mount,
mounts,
service_start,
service_stop,
service_running,
umount,
cmp_pkgrevno,
)
from charmhelpers.fetch import (
apt_install,
)
from charmhelpers.core.unitdata import kv
from charmhelpers.core.kernel import modprobe
from charmhelpers.contrib.openstack.utils import config_flags_parser
KEYRING = '/etc/ceph/ceph.client.{}.keyring'
KEYFILE = '/etc/ceph/ceph.client.{}.key'
CEPH_CONF = """[global]
auth supported = {auth}
keyring = {keyring}
mon host = {mon_hosts}
log to syslog = {use_syslog}
err to syslog = {use_syslog}
clog to syslog = {use_syslog}
"""
# The number of placement groups per OSD to target for placement group
# calculations. This number is chosen as 100 due to the ceph PG Calc
# documentation recommending to choose 100 for clusters which are not
# expected to increase in the foreseeable future. Since the majority of the
# calculations are done on deployment, target the case of non-expanding
# clusters as the default.
DEFAULT_PGS_PER_OSD_TARGET = 100
DEFAULT_POOL_WEIGHT = 10.0
LEGACY_PG_COUNT = 200
DEFAULT_MINIMUM_PGS = 2
AUTOSCALER_DEFAULT_PGS = 32
class OsdPostUpgradeError(Exception):
"""Error class for OSD post-upgrade operations."""
pass
class OSDSettingConflict(Exception):
"""Error class for conflicting osd setting requests."""
pass
class OSDSettingNotAllowed(Exception):
"""Error class for a disallowed setting."""
pass
OSD_SETTING_EXCEPTIONS = (OSDSettingConflict, OSDSettingNotAllowed)
OSD_SETTING_WHITELIST = [
'osd heartbeat grace',
'osd heartbeat interval',
]
def _order_dict_by_key(rdict):
"""Convert a dictionary into an OrderedDict sorted by key.
:param rdict: Dictionary to be ordered.
:type rdict: dict
:returns: Ordered Dictionary.
:rtype: collections.OrderedDict
"""
return collections.OrderedDict(sorted(rdict.items(), key=lambda k: k[0]))
def get_osd_settings(relation_name):
"""Consolidate requested osd settings from all clients.
Consolidate requested osd settings from all clients. Check that the
requested setting is on the whitelist and it does not conflict with
any other requested settings.
:returns: Dictionary of settings
:rtype: dict
:raises: OSDSettingNotAllowed
:raises: OSDSettingConflict
"""
rel_ids = relation_ids(relation_name)
osd_settings = {}
for relid in rel_ids:
for unit in related_units(relid):
unit_settings = relation_get('osd-settings', unit, relid) or '{}'
unit_settings = json.loads(unit_settings)
for key, value in unit_settings.items():
if key not in OSD_SETTING_WHITELIST:
msg = 'Illegal settings "{}"'.format(key)
raise OSDSettingNotAllowed(msg)
if key in osd_settings:
if osd_settings[key] != unit_settings[key]:
msg = 'Conflicting settings for "{}"'.format(key)
raise OSDSettingConflict(msg)
else:
osd_settings[key] = value
return _order_dict_by_key(osd_settings)
def send_osd_settings():
"""Pass on requested OSD settings to osd units."""
try:
settings = get_osd_settings('client')
except OSD_SETTING_EXCEPTIONS as e:
# There is a problem with the settings, not passing them on. Update
# status will notify the user.
log(e, level=ERROR)
return
data = {
'osd-settings': json.dumps(settings, sort_keys=True)}
for relid in relation_ids('osd'):
relation_set(relation_id=relid,
relation_settings=data)
def validator(value, valid_type, valid_range=None):
"""
Used to validate these: http://docs.ceph.com/docs/master/rados/operations/pools/#set-pool-values
Example input:
validator(value=1,
valid_type=int,
valid_range=[0, 2])
This says I'm testing value=1. It must be an int inclusive in [0,2]
:param value: The value to validate
:param valid_type: The type that value should be.
:param valid_range: A range of values that value can assume.
:return:
"""
assert isinstance(value, valid_type), "{} is not a {}".format(
value,
valid_type)
if valid_range is not None:
assert isinstance(valid_range, list), \
"valid_range must be a list, was given {}".format(valid_range)
# If we're dealing with strings
if isinstance(value, six.string_types):
assert value in valid_range, \
"{} is not in the list {}".format(value, valid_range)
# Integer, float should have a min and max
else:
if len(valid_range) != 2:
raise ValueError(
"Invalid valid_range list of {} for {}. "
"List must be [min,max]".format(valid_range, value))
assert value >= valid_range[0], \
"{} is less than minimum allowed value of {}".format(
value, valid_range[0])
assert value <= valid_range[1], \
"{} is greater than maximum allowed value of {}".format(
value, valid_range[1])
class PoolCreationError(Exception):
"""
A custom error to inform the caller that a pool creation failed. Provides an error message
"""
def __init__(self, message):
super(PoolCreationError, self).__init__(message)
class Pool(object):
"""
An object oriented approach to Ceph pool creation. This base class is inherited by ReplicatedPool and ErasurePool.
Do not call create() on this base class as it will not do anything. Instantiate a child class and call create().
"""
def __init__(self, service, name):
self.service = service
self.name = name
# Create the pool if it doesn't exist already
# To be implemented by subclasses
def create(self):
pass
def add_cache_tier(self, cache_pool, mode):
"""
Adds a new cache tier to an existing pool.
:param cache_pool: six.string_types. The cache tier pool name to add.
:param mode: six.string_types. The caching mode to use for this pool. valid range = ["readonly", "writeback"]
:return: None
"""
# Check the input types and values
validator(value=cache_pool, valid_type=six.string_types)
validator(value=mode, valid_type=six.string_types, valid_range=["readonly", "writeback"])
check_call(['ceph', '--id', self.service, 'osd', 'tier', 'add', self.name, cache_pool])
check_call(['ceph', '--id', self.service, 'osd', 'tier', 'cache-mode', cache_pool, mode])
check_call(['ceph', '--id', self.service, 'osd', 'tier', 'set-overlay', self.name, cache_pool])
check_call(['ceph', '--id', self.service, 'osd', 'pool', 'set', cache_pool, 'hit_set_type', 'bloom'])
def remove_cache_tier(self, cache_pool):
"""
Removes a cache tier from Ceph. Flushes all dirty objects from writeback pools and waits for that to complete.
:param cache_pool: six.string_types. The cache tier pool name to remove.
:return: None
"""
# read-only is easy, writeback is much harder
mode = get_cache_mode(self.service, cache_pool)
if mode == 'readonly':
check_call(['ceph', '--id', self.service, 'osd', 'tier', 'cache-mode', cache_pool, 'none'])
check_call(['ceph', '--id', self.service, 'osd', 'tier', 'remove', self.name, cache_pool])
elif mode == 'writeback':
pool_forward_cmd = ['ceph', '--id', self.service, 'osd', 'tier',
'cache-mode', cache_pool, 'forward']
if cmp_pkgrevno('ceph-common', '10.1') >= 0:
# Jewel added a mandatory flag
pool_forward_cmd.append('--yes-i-really-mean-it')
check_call(pool_forward_cmd)
# Flush the cache and wait for it to return
check_call(['rados', '--id', self.service, '-p', cache_pool, 'cache-flush-evict-all'])
check_call(['ceph', '--id', self.service, 'osd', 'tier', 'remove-overlay', self.name])
check_call(['ceph', '--id', self.service, 'osd', 'tier', 'remove', self.name, cache_pool])
def get_pgs(self, pool_size, percent_data=DEFAULT_POOL_WEIGHT,
device_class=None):
"""Return the number of placement groups to use when creating the pool.
Returns the number of placement groups which should be specified when
creating the pool. This is based upon the calculation guidelines
provided by the Ceph Placement Group Calculator (located online at
http://ceph.com/pgcalc/).
The number of placement groups are calculated using the following:
(Target PGs per OSD) * (OSD #) * (%Data)
----------------------------------------
(Pool size)
Per the upstream guidelines, the OSD # should really be considered
based on the number of OSDs which are eligible to be selected by the
pool. Since the pool creation doesn't specify any of CRUSH set rules,
the default rule will be dependent upon the type of pool being
created (replicated or erasure).
This code makes no attempt to determine the number of OSDs which can be
selected for the specific rule, rather it is left to the user to tune
in the form of 'expected-osd-count' config option.
:param pool_size: int. pool_size is either the number of replicas for
replicated pools or the K+M sum for erasure coded pools
:param percent_data: float. the percentage of data that is expected to
be contained in the pool for the specific OSD set. Default value
is to assume 10% of the data is for this pool, which is a
relatively low % of the data but allows for the pg_num to be
increased. NOTE: the default is primarily to handle the scenario
where related charms requiring pools has not been upgraded to
include an update to indicate their relative usage of the pools.
:param device_class: str. class of storage to use for basis of pgs
calculation; ceph supports nvme, ssd and hdd by default based
on presence of devices of each type in the deployment.
:return: int. The number of pgs to use.
"""
# Note: This calculation follows the approach that is provided
# by the Ceph PG Calculator located at http://ceph.com/pgcalc/.
validator(value=pool_size, valid_type=int)
# Ensure that percent data is set to something - even with a default
# it can be set to None, which would wreak havoc below.
if percent_data is None:
percent_data = DEFAULT_POOL_WEIGHT
# If the expected-osd-count is specified, then use the max between
# the expected-osd-count and the actual osd_count
osd_list = get_osds(self.service, device_class)
expected = config('expected-osd-count') or 0
if osd_list:
if device_class:
osd_count = len(osd_list)
else:
osd_count = max(expected, len(osd_list))
# Log a message to provide some insight if the calculations claim
# to be off because someone is setting the expected count and
# there are more OSDs in reality. Try to make a proper guess
# based upon the cluster itself.
if not device_class and expected and osd_count != expected:
log("Found more OSDs than provided expected count. "
"Using the actual count instead", INFO)
elif expected:
# Use the expected-osd-count in older ceph versions to allow for
# a more accurate pg calculations
osd_count = expected
else:
# NOTE(james-page): Default to 200 for older ceph versions
# which don't support OSD query from cli
return LEGACY_PG_COUNT
percent_data /= 100.0
target_pgs_per_osd = config('pgs-per-osd') or DEFAULT_PGS_PER_OSD_TARGET
num_pg = (target_pgs_per_osd * osd_count * percent_data) // pool_size
# NOTE: ensure a sane minimum number of PGS otherwise we don't get any
# reasonable data distribution in minimal OSD configurations
if num_pg < DEFAULT_MINIMUM_PGS:
num_pg = DEFAULT_MINIMUM_PGS
# The CRUSH algorithm has a slight optimization for placement groups
# with powers of 2 so find the nearest power of 2. If the nearest
# power of 2 is more than 25% below the original value, the next
# highest value is used. To do this, find the nearest power of 2 such
# that 2^n <= num_pg, check to see if its within the 25% tolerance.
exponent = math.floor(math.log(num_pg, 2))
nearest = 2 ** exponent
if (num_pg - nearest) > (num_pg * 0.25):
# Choose the next highest power of 2 since the nearest is more
# than 25% below the original value.
return int(nearest * 2)
else:
return int(nearest)
class ReplicatedPool(Pool):
def __init__(self, service, name, pg_num=None, replicas=2,
percent_data=10.0, app_name=None):
super(ReplicatedPool, self).__init__(service=service, name=name)
self.replicas = replicas
self.percent_data = percent_data
if pg_num:
# Since the number of placement groups were specified, ensure
# that there aren't too many created.
max_pgs = self.get_pgs(self.replicas, 100.0)
self.pg_num = min(pg_num, max_pgs)
else:
self.pg_num = self.get_pgs(self.replicas, percent_data)
if app_name:
self.app_name = app_name
else:
self.app_name = 'unknown'
def create(self):
if not pool_exists(self.service, self.name):
nautilus_or_later = cmp_pkgrevno('ceph-common', '14.2.0') >= 0
# Create it
if nautilus_or_later:
cmd = [
'ceph', '--id', self.service, 'osd', 'pool', 'create',
'--pg-num-min={}'.format(
min(AUTOSCALER_DEFAULT_PGS, self.pg_num)
),
self.name, str(self.pg_num)
]
else:
cmd = [
'ceph', '--id', self.service, 'osd', 'pool', 'create',
self.name, str(self.pg_num)
]
try:
check_call(cmd)
# Set the pool replica size
update_pool(client=self.service,
pool=self.name,
settings={'size': str(self.replicas)})
if nautilus_or_later:
# Ensure we set the expected pool ratio
update_pool(client=self.service,
pool=self.name,
settings={'target_size_ratio': str(self.percent_data / 100.0)})
try:
set_app_name_for_pool(client=self.service,
pool=self.name,
name=self.app_name)
except CalledProcessError:
log('Could not set app name for pool {}'.format(self.name), level=WARNING)
if 'pg_autoscaler' in enabled_manager_modules():
try:
enable_pg_autoscale(self.service, self.name)
except CalledProcessError as e:
log('Could not configure auto scaling for pool {}: {}'.format(
self.name, e), level=WARNING)
except CalledProcessError:
raise
# Default jerasure erasure coded pool
class ErasurePool(Pool):
def __init__(self, service, name, erasure_code_profile="default",
percent_data=10.0, app_name=None):
super(ErasurePool, self).__init__(service=service, name=name)
self.erasure_code_profile = erasure_code_profile
self.percent_data = percent_data
if app_name:
self.app_name = app_name
else:
self.app_name = 'unknown'
def create(self):
if not pool_exists(self.service, self.name):
# Try to find the erasure profile information in order to properly
# size the number of placement groups. The size of an erasure
# coded placement group is calculated as k+m.
erasure_profile = get_erasure_profile(self.service,
self.erasure_code_profile)
# Check for errors
if erasure_profile is None:
msg = ("Failed to discover erasure profile named "
"{}".format(self.erasure_code_profile))
log(msg, level=ERROR)
raise PoolCreationError(msg)
if 'k' not in erasure_profile or 'm' not in erasure_profile:
# Error
msg = ("Unable to find k (data chunks) or m (coding chunks) "
"in erasure profile {}".format(erasure_profile))
log(msg, level=ERROR)
raise PoolCreationError(msg)
k = int(erasure_profile['k'])
m = int(erasure_profile['m'])
pgs = self.get_pgs(k + m, self.percent_data)
nautilus_or_later = cmp_pkgrevno('ceph-common', '14.2.0') >= 0
# Create it
if nautilus_or_later:
cmd = [
'ceph', '--id', self.service, 'osd', 'pool', 'create',
'--pg-num-min={}'.format(
min(AUTOSCALER_DEFAULT_PGS, pgs)
),
self.name, str(pgs), str(pgs),
'erasure', self.erasure_code_profile
]
else:
cmd = [
'ceph', '--id', self.service, 'osd', 'pool', 'create',
self.name, str(pgs), str(pgs),
'erasure', self.erasure_code_profile
]
try:
check_call(cmd)
try:
set_app_name_for_pool(client=self.service,
pool=self.name,
name=self.app_name)
except CalledProcessError:
log('Could not set app name for pool {}'.format(self.name), level=WARNING)
if nautilus_or_later:
# Ensure we set the expected pool ratio
update_pool(client=self.service,
pool=self.name,
settings={'target_size_ratio': str(self.percent_data / 100.0)})
if 'pg_autoscaler' in enabled_manager_modules():
try:
enable_pg_autoscale(self.service, self.name)
except CalledProcessError as e:
log('Could not configure auto scaling for pool {}: {}'.format(
self.name, e), level=WARNING)
except CalledProcessError:
raise
"""Get an existing erasure code profile if it already exists.
Returns json formatted output"""
def enabled_manager_modules():
"""Return a list of enabled manager modules.
:rtype: List[str]
"""
cmd = ['ceph', 'mgr', 'module', 'ls']
try:
modules = check_output(cmd)
if six.PY3:
modules = modules.decode('UTF-8')
except CalledProcessError as e:
log("Failed to list ceph modules: {}".format(e), WARNING)
return []
modules = json.loads(modules)
return modules['enabled_modules']
def enable_pg_autoscale(service, pool_name):
"""
Enable Ceph's PG autoscaler for the specified pool.
:param service: six.string_types. The Ceph user name to run the command under
:param pool_name: six.string_types. The name of the pool to enable sutoscaling on
:raise: CalledProcessError if the command fails
"""
check_call(['ceph', '--id', service, 'osd', 'pool', 'set', pool_name, 'pg_autoscale_mode', 'on'])
def get_mon_map(service):
"""
Returns the current monitor map.
:param service: six.string_types. The Ceph user name to run the command under
:return: json string. :raise: ValueError if the monmap fails to parse.
Also raises CalledProcessError if our ceph command fails
"""
try:
mon_status = check_output(['ceph', '--id', service,
'mon_status', '--format=json'])
if six.PY3:
mon_status = mon_status.decode('UTF-8')
try:
return json.loads(mon_status)
except ValueError as v:
log("Unable to parse mon_status json: {}. Error: {}"
.format(mon_status, str(v)))
raise
except CalledProcessError as e:
log("mon_status command failed with message: {}"
.format(str(e)))
raise
def hash_monitor_names(service):
"""
Uses the get_mon_map() function to get information about the monitor
cluster.
Hash the name of each monitor. Return a sorted list of monitor hashes
in an ascending order.
:param service: six.string_types. The Ceph user name to run the command under
:rtype : dict. json dict of monitor name, ip address and rank
example: {
'name': 'ip-172-31-13-165',
'rank': 0,
'addr': '172.31.13.165:6789/0'}
"""
try:
hash_list = []
monitor_list = get_mon_map(service=service)
if monitor_list['monmap']['mons']:
for mon in monitor_list['monmap']['mons']:
hash_list.append(
hashlib.sha224(mon['name'].encode('utf-8')).hexdigest())
return sorted(hash_list)
else:
return None
except (ValueError, CalledProcessError):
raise
def monitor_key_delete(service, key):
"""
Delete a key and value pair from the monitor cluster
:param service: six.string_types. The Ceph user name to run the command under
Deletes a key value pair on the monitor cluster.
:param key: six.string_types. The key to delete.
"""
try:
check_output(
['ceph', '--id', service,
'config-key', 'del', str(key)])
except CalledProcessError as e:
log("Monitor config-key put failed with message: {}".format(
e.output))
raise
def monitor_key_set(service, key, value):
"""
Sets a key value pair on the monitor cluster.
:param service: six.string_types. The Ceph user name to run the command under
:param key: six.string_types. The key to set.
:param value: The value to set. This will be converted to a string
before setting
"""
try:
check_output(
['ceph', '--id', service,
'config-key', 'put', str(key), str(value)])
except CalledProcessError as e:
log("Monitor config-key put failed with message: {}".format(
e.output))
raise
def monitor_key_get(service, key):
"""
Gets the value of an existing key in the monitor cluster.
:param service: six.string_types. The Ceph user name to run the command under
:param key: six.string_types. The key to search for.
:return: Returns the value of that key or None if not found.
"""
try:
output = check_output(
['ceph', '--id', service,
'config-key', 'get', str(key)]).decode('UTF-8')
return output
except CalledProcessError as e:
log("Monitor config-key get failed with message: {}".format(
e.output))
return None
def monitor_key_exists(service, key):
"""
Searches for the existence of a key in the monitor cluster.
:param service: six.string_types. The Ceph user name to run the command under
:param key: six.string_types. The key to search for
:return: Returns True if the key exists, False if not and raises an
exception if an unknown error occurs. :raise: CalledProcessError if
an unknown error occurs
"""
try:
check_call(
['ceph', '--id', service,
'config-key', 'exists', str(key)])
# I can return true here regardless because Ceph returns
# ENOENT if the key wasn't found
return True
except CalledProcessError as e:
if e.returncode == errno.ENOENT:
return False
else:
log("Unknown error from ceph config-get exists: {} {}".format(
e.returncode, e.output))
raise
def get_erasure_profile(service, name):
"""
:param service: six.string_types. The Ceph user name to run the command under
:param name:
:return:
"""
try:
out = check_output(['ceph', '--id', service,
'osd', 'erasure-code-profile', 'get',
name, '--format=json'])
if six.PY3:
out = out.decode('UTF-8')
return json.loads(out)
except (CalledProcessError, OSError, ValueError):
return None
def pool_set(service, pool_name, key, value):
"""
Sets a value for a RADOS pool in ceph.
:param service: six.string_types. The Ceph user name to run the command under
:param pool_name: six.string_types
:param key: six.string_types
:param value:
:return: None. Can raise CalledProcessError
"""
cmd = ['ceph', '--id', service, 'osd', 'pool', 'set', pool_name, key,
str(value).lower()]
try:
check_call(cmd)
except CalledProcessError:
raise
def snapshot_pool(service, pool_name, snapshot_name):
"""
Snapshots a RADOS pool in ceph.
:param service: six.string_types. The Ceph user name to run the command under
:param pool_name: six.string_types
:param snapshot_name: six.string_types
:return: None. Can raise CalledProcessError
"""
cmd = ['ceph', '--id', service, 'osd', 'pool', 'mksnap', pool_name, snapshot_name]
try:
check_call(cmd)
except CalledProcessError:
raise
def remove_pool_snapshot(service, pool_name, snapshot_name):
"""
Remove a snapshot from a RADOS pool in ceph.
:param service: six.string_types. The Ceph user name to run the command under
:param pool_name: six.string_types
:param snapshot_name: six.string_types
:return: None. Can raise CalledProcessError
"""
cmd = ['ceph', '--id', service, 'osd', 'pool', 'rmsnap', pool_name, snapshot_name]
try:
check_call(cmd)
except CalledProcessError:
raise
def set_pool_quota(service, pool_name, max_bytes=None, max_objects=None):
"""
:param service: The Ceph user name to run the command under
:type service: str
:param pool_name: Name of pool
:type pool_name: str
:param max_bytes: Maximum bytes quota to apply
:type max_bytes: int
:param max_objects: Maximum objects quota to apply
:type max_objects: int
:raises: subprocess.CalledProcessError
"""
cmd = ['ceph', '--id', service, 'osd', 'pool', 'set-quota', pool_name]
if max_bytes:
cmd = cmd + ['max_bytes', str(max_bytes)]
if max_objects:
cmd = cmd + ['max_objects', str(max_objects)]
check_call(cmd)
def remove_pool_quota(service, pool_name):
"""
Set a byte quota on a RADOS pool in ceph.
:param service: six.string_types. The Ceph user name to run the command under
:param pool_name: six.string_types
:return: None. Can raise CalledProcessError
"""
cmd = ['ceph', '--id', service, 'osd', 'pool', 'set-quota', pool_name, 'max_bytes', '0']
try:
check_call(cmd)
except CalledProcessError:
raise
def remove_erasure_profile(service, profile_name):
"""
Create a new erasure code profile if one does not already exist for it. Updates
the profile if it exists. Please see http://docs.ceph.com/docs/master/rados/operations/erasure-code-profile/
for more details
:param service: six.string_types. The Ceph user name to run the command under
:param profile_name: six.string_types
:return: None. Can raise CalledProcessError
"""
cmd = ['ceph', '--id', service, 'osd', 'erasure-code-profile', 'rm',
profile_name]
try:
check_call(cmd)
except CalledProcessError:
raise
def create_erasure_profile(service, profile_name, erasure_plugin_name='jerasure',
failure_domain='host',
data_chunks=2, coding_chunks=1,
locality=None, durability_estimator=None,
device_class=None):
"""
Create a new erasure code profile if one does not already exist for it. Updates
the profile if it exists. Please see http://docs.ceph.com/docs/master/rados/operations/erasure-code-profile/
for more details
:param service: six.string_types. The Ceph user name to run the command under
:param profile_name: six.string_types
:param erasure_plugin_name: six.string_types
:param failure_domain: six.string_types. One of ['chassis', 'datacenter', 'host', 'osd', 'pdu', 'pod', 'rack', 'region',
'room', 'root', 'row'])
:param data_chunks: int
:param coding_chunks: int
:param locality: int
:param durability_estimator: int
:param device_class: six.string_types
:return: None. Can raise CalledProcessError
"""
# Ensure this failure_domain is allowed by Ceph
validator(failure_domain, six.string_types,
['chassis', 'datacenter', 'host', 'osd', 'pdu', 'pod', 'rack', 'region', 'room', 'root', 'row'])
cmd = ['ceph', '--id', service, 'osd', 'erasure-code-profile', 'set', profile_name,
'plugin=' + erasure_plugin_name, 'k=' + str(data_chunks), 'm=' + str(coding_chunks)
]
if locality is not None and durability_estimator is not None:
raise ValueError("create_erasure_profile should be called with k, m and one of l or c but not both.")
luminous_or_later = cmp_pkgrevno('ceph-common', '12.0.0') >= 0
# failure_domain changed in luminous
if luminous_or_later:
cmd.append('crush-failure-domain=' + failure_domain)
else:
cmd.append('ruleset-failure-domain=' + failure_domain)
# device class new in luminous
if luminous_or_later and device_class:
cmd.append('crush-device-class={}'.format(device_class))
else:
log('Skipping device class configuration (ceph < 12.0.0)',
level=DEBUG)
# Add plugin specific information
if locality is not None:
# For local erasure codes
cmd.append('l=' + str(locality))
if durability_estimator is not None:
# For Shec erasure codes
cmd.append('c=' + str(durability_estimator))
if erasure_profile_exists(service, profile_name):
cmd.append('--force')
try:
check_call(cmd)
except CalledProcessError:
raise
def rename_pool(service, old_name, new_name):
"""
Rename a Ceph pool from old_name to new_name
:param service: six.string_types. The Ceph user name to run the command under
:param old_name: six.string_types
:param new_name: six.string_types
:return: None
"""
validator(value=old_name, valid_type=six.string_types)
validator(value=new_name, valid_type=six.string_types)
cmd = ['ceph', '--id', service, 'osd', 'pool', 'rename', old_name, new_name]
check_call(cmd)
def erasure_profile_exists(service, name):
"""
Check to see if an Erasure code profile already exists.
:param service: six.string_types. The Ceph user name to run the command under
:param name: six.string_types
:return: int or None
"""
validator(value=name, valid_type=six.string_types)
try:
check_call(['ceph', '--id', service,
'osd', 'erasure-code-profile', 'get',
name])
return True
except CalledProcessError:
return False
def get_cache_mode(service, pool_name):
"""
Find the current caching mode of the pool_name given.
:param service: six.string_types. The Ceph user name to run the command under
:param pool_name: six.string_types
:return: int or None
"""
validator(value=service, valid_type=six.string_types)
validator(value=pool_name, valid_type=six.string_types)
out = check_output(['ceph', '--id', service,
'osd', 'dump', '--format=json'])
if six.PY3:
out = out.decode('UTF-8')
try:
osd_json = json.loads(out)
for pool in osd_json['pools']:
if pool['pool_name'] == pool_name:
return pool['cache_mode']
return None
except ValueError:
raise
def pool_exists(service, name):
"""Check to see if a RADOS pool already exists."""
try:
out = check_output(['rados', '--id', service, 'lspools'])
if six.PY3:
out = out.decode('UTF-8')
except CalledProcessError:
return False
return name in out.split()
def get_osds(service, device_class=None):
"""Return a list of all Ceph Object Storage Daemons currently in the
cluster (optionally filtered by storage device class).
:param device_class: Class of storage device for OSD's
:type device_class: str
"""
luminous_or_later = cmp_pkgrevno('ceph-common', '12.0.0') >= 0
if luminous_or_later and device_class:
out = check_output(['ceph', '--id', service,
'osd', 'crush', 'class',
'ls-osd', device_class,
'--format=json'])
else:
out = check_output(['ceph', '--id', service,
'osd', 'ls',
'--format=json'])
if six.PY3:
out = out.decode('UTF-8')
return json.loads(out)
def install():
"""Basic Ceph client installation."""
ceph_dir = "/etc/ceph"
if not os.path.exists(ceph_dir):
os.mkdir(ceph_dir)
apt_install('ceph-common', fatal=True)
def rbd_exists(service, pool, rbd_img):
"""Check to see if a RADOS block device exists."""
try:
out = check_output(['rbd', 'list', '--id',
service, '--pool', pool])
if six.PY3:
out = out.decode('UTF-8')
except CalledProcessError:
return False
return rbd_img in out
def create_rbd_image(service, pool, image, sizemb):
"""Create a new RADOS block device."""
cmd = ['rbd', 'create', image, '--size', str(sizemb), '--id', service,
'--pool', pool]
check_call(cmd)
def update_pool(client, pool, settings):
cmd = ['ceph', '--id', client, 'osd', 'pool', 'set', pool]
for k, v in six.iteritems(settings):
cmd.append(k)
cmd.append(v)
check_call(cmd)
def set_app_name_for_pool(client, pool, name):
"""
Calls `osd pool application enable` for the specified pool name
:param client: Name of the ceph client to use
:type client: str
:param pool: Pool to set app name for
:type pool: str
:param name: app name for the specified pool
:type name: str
:raises: CalledProcessError if ceph call fails
"""
if cmp_pkgrevno('ceph-common', '12.0.0') >= 0: