-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathosd-scrub-repair.sh
executable file
·6212 lines (5920 loc) · 167 KB
/
osd-scrub-repair.sh
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
#!/usr/bin/env bash
#
# Copyright (C) 2014 Red Hat <[email protected]>
#
# Author: Loic Dachary <[email protected]>
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU Library Public License as published by
# the Free Software Foundation; either version 2, or (at your option)
# any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU Library Public License for more details.
#
set -x
source $CEPH_ROOT/qa/standalone/ceph-helpers.sh
if [ `uname` = FreeBSD ]; then
# erasure coding overwrites are only tested on Bluestore
# erasure coding on filestore is unsafe
# http://docs.ceph.com/en/latest/rados/operations/erasure-code/#erasure-coding-with-overwrites
use_ec_overwrite=false
else
use_ec_overwrite=true
fi
# Test development and debugging
# Set to "yes" in order to ignore diff errors and save results to update test
getjson="no"
# Filter out mtime and local_mtime dates, version, prior_version and last_reqid (client) from any object_info.
jqfilter='def walk(f):
. as $in
| if type == "object" then
reduce keys[] as $key
( {}; . + { ($key): ($in[$key] | walk(f)) } ) | f
elif type == "array" then map( walk(f) ) | f
else f
end;
walk(if type == "object" then del(.mtime) else . end)
| walk(if type == "object" then del(.local_mtime) else . end)
| walk(if type == "object" then del(.last_reqid) else . end)
| walk(if type == "object" then del(.version) else . end)
| walk(if type == "object" then del(.prior_version) else . end)'
sortkeys='import json; import sys ; JSON=sys.stdin.read() ; ud = json.loads(JSON) ; print(json.dumps(ud, sort_keys=True, indent=2))'
function run() {
local dir=$1
shift
export CEPH_MON="127.0.0.1:7107" # git grep '\<7107\>' : there must be only one
export CEPH_ARGS
CEPH_ARGS+="--fsid=$(uuidgen) --auth-supported=none "
CEPH_ARGS+="--mon-host=$CEPH_MON "
CEPH_ARGS+="--osd-skip-data-digest=false "
export -n CEPH_CLI_TEST_DUP_COMMAND
local funcs=${@:-$(set | sed -n -e 's/^\(TEST_[0-9a-z_]*\) .*/\1/p')}
for func in $funcs ; do
$func $dir || return 1
done
}
function add_something() {
local dir=$1
local poolname=$2
local obj=${3:-SOMETHING}
local scrub=${4:-noscrub}
if [ "$scrub" = "noscrub" ];
then
ceph osd set noscrub || return 1
ceph osd set nodeep-scrub || return 1
else
ceph osd unset noscrub || return 1
ceph osd unset nodeep-scrub || return 1
fi
local payload=ABCDEF
echo $payload > $dir/ORIGINAL
rados --pool $poolname put $obj $dir/ORIGINAL || return 1
}
#
# Corrupt one copy of a replicated pool
#
function TEST_corrupt_and_repair_replicated() {
local dir=$1
local poolname=rbd
setup $dir || return 1
run_mon $dir a --osd_pool_default_size=2 || return 1
run_mgr $dir x || return 1
run_osd $dir 0 || return 1
run_osd $dir 1 || return 1
create_rbd_pool || return 1
wait_for_clean || return 1
add_something $dir $poolname || return 1
corrupt_and_repair_one $dir $poolname $(get_not_primary $poolname SOMETHING) || return 1
# Reproduces http://tracker.ceph.com/issues/8914
corrupt_and_repair_one $dir $poolname $(get_primary $poolname SOMETHING) || return 1
teardown $dir || return 1
}
#
# Allow repair to be scheduled when some recovering is still undergoing on the same OSD
#
function TEST_allow_repair_during_recovery() {
local dir=$1
local poolname=rbd
setup $dir || return 1
run_mon $dir a --osd_pool_default_size=2 || return 1
run_mgr $dir x || return 1
run_osd $dir 0 --osd_scrub_during_recovery=false \
--osd_repair_during_recovery=true \
--osd_debug_pretend_recovery_active=true || return 1
run_osd $dir 1 --osd_scrub_during_recovery=false \
--osd_repair_during_recovery=true \
--osd_debug_pretend_recovery_active=true || return 1
create_rbd_pool || return 1
wait_for_clean || return 1
add_something $dir $poolname || return 1
corrupt_and_repair_one $dir $poolname $(get_not_primary $poolname SOMETHING) || return 1
teardown $dir || return 1
}
#
# Skip non-repair scrub correctly during recovery
#
function TEST_skip_non_repair_during_recovery() {
local dir=$1
local poolname=rbd
setup $dir || return 1
run_mon $dir a --osd_pool_default_size=2 || return 1
run_mgr $dir x || return 1
run_osd $dir 0 --osd_scrub_during_recovery=false \
--osd_repair_during_recovery=true \
--osd_debug_pretend_recovery_active=true || return 1
run_osd $dir 1 --osd_scrub_during_recovery=false \
--osd_repair_during_recovery=true \
--osd_debug_pretend_recovery_active=true || return 1
create_rbd_pool || return 1
wait_for_clean || return 1
add_something $dir $poolname || return 1
scrub_and_not_schedule $dir $poolname $(get_not_primary $poolname SOMETHING) || return 1
teardown $dir || return 1
}
function scrub_and_not_schedule() {
local dir=$1
local poolname=$2
local osd=$3
#
# 1) start a non-repair scrub
#
local pg=$(get_pg $poolname SOMETHING)
local last_scrub=$(get_last_scrub_stamp $pg)
ceph pg scrub $pg
#
# 2) Assure the scrub is not scheduled
#
for ((i=0; i < 3; i++)); do
if test "$(get_last_scrub_stamp $pg)" '>' "$last_scrub" ; then
return 1
fi
sleep 1
done
#
# 3) Access to the file must OK
#
objectstore_tool $dir $osd SOMETHING list-attrs || return 1
rados --pool $poolname get SOMETHING $dir/COPY || return 1
diff $dir/ORIGINAL $dir/COPY || return 1
}
function corrupt_and_repair_two() {
local dir=$1
local poolname=$2
local first=$3
local second=$4
#
# 1) remove the corresponding file from the OSDs
#
pids=""
run_in_background pids objectstore_tool $dir $first SOMETHING remove
run_in_background pids objectstore_tool $dir $second SOMETHING remove
wait_background pids
return_code=$?
if [ $return_code -ne 0 ]; then return $return_code; fi
#
# 2) repair the PG
#
local pg=$(get_pg $poolname SOMETHING)
repair $pg
#
# 3) The files must be back
#
pids=""
run_in_background pids objectstore_tool $dir $first SOMETHING list-attrs
run_in_background pids objectstore_tool $dir $second SOMETHING list-attrs
wait_background pids
return_code=$?
if [ $return_code -ne 0 ]; then return $return_code; fi
rados --pool $poolname get SOMETHING $dir/COPY || return 1
diff $dir/ORIGINAL $dir/COPY || return 1
}
#
# 1) add an object
# 2) remove the corresponding file from a designated OSD
# 3) repair the PG
# 4) check that the file has been restored in the designated OSD
#
function corrupt_and_repair_one() {
local dir=$1
local poolname=$2
local osd=$3
#
# 1) remove the corresponding file from the OSD
#
objectstore_tool $dir $osd SOMETHING remove || return 1
#
# 2) repair the PG
#
local pg=$(get_pg $poolname SOMETHING)
repair $pg
#
# 3) The file must be back
#
objectstore_tool $dir $osd SOMETHING list-attrs || return 1
rados --pool $poolname get SOMETHING $dir/COPY || return 1
diff $dir/ORIGINAL $dir/COPY || return 1
}
function corrupt_and_repair_erasure_coded() {
local dir=$1
local poolname=$2
add_something $dir $poolname || return 1
local primary=$(get_primary $poolname SOMETHING)
local -a osds=($(get_osds $poolname SOMETHING | sed -e "s/$primary//"))
local not_primary_first=${osds[0]}
local not_primary_second=${osds[1]}
# Reproduces http://tracker.ceph.com/issues/10017
corrupt_and_repair_one $dir $poolname $primary || return 1
# Reproduces http://tracker.ceph.com/issues/10409
corrupt_and_repair_one $dir $poolname $not_primary_first || return 1
corrupt_and_repair_two $dir $poolname $not_primary_first $not_primary_second || return 1
corrupt_and_repair_two $dir $poolname $primary $not_primary_first || return 1
}
function auto_repair_erasure_coded() {
local dir=$1
local allow_overwrites=$2
local poolname=ecpool
# Launch a cluster with 5 seconds scrub interval
setup $dir || return 1
run_mon $dir a || return 1
run_mgr $dir x || return 1
local ceph_osd_args="--osd-scrub-auto-repair=true \
--osd-deep-scrub-interval=5 \
--osd-scrub-max-interval=5 \
--osd-scrub-min-interval=5 \
--osd-scrub-interval-randomize-ratio=0"
for id in $(seq 0 2) ; do
if [ "$allow_overwrites" = "true" ]; then
run_osd $dir $id $ceph_osd_args || return 1
else
run_osd_filestore $dir $id $ceph_osd_args || return 1
fi
done
create_rbd_pool || return 1
wait_for_clean || return 1
# Create an EC pool
create_ec_pool $poolname $allow_overwrites k=2 m=1 || return 1
# Put an object
local payload=ABCDEF
echo $payload > $dir/ORIGINAL
rados --pool $poolname put SOMETHING $dir/ORIGINAL || return 1
# Remove the object from one shard physically
# Restarted osd get $ceph_osd_args passed
objectstore_tool $dir $(get_not_primary $poolname SOMETHING) SOMETHING remove || return 1
# Wait for auto repair
local pgid=$(get_pg $poolname SOMETHING)
wait_for_scrub $pgid "$(get_last_scrub_stamp $pgid)"
wait_for_clean || return 1
# Verify - the file should be back
# Restarted osd get $ceph_osd_args passed
objectstore_tool $dir $(get_not_primary $poolname SOMETHING) SOMETHING list-attrs || return 1
rados --pool $poolname get SOMETHING $dir/COPY || return 1
diff $dir/ORIGINAL $dir/COPY || return 1
# Tear down
teardown $dir || return 1
}
function TEST_auto_repair_erasure_coded_appends() {
auto_repair_erasure_coded $1 false
}
function TEST_auto_repair_erasure_coded_overwrites() {
if [ "$use_ec_overwrite" = "true" ]; then
auto_repair_erasure_coded $1 true
fi
}
function TEST_auto_repair_bluestore_basic() {
local dir=$1
local poolname=testpool
# Launch a cluster with 5 seconds scrub interval
setup $dir || return 1
run_mon $dir a || return 1
run_mgr $dir x || return 1
local ceph_osd_args="--osd-scrub-auto-repair=true \
--osd_deep_scrub_randomize_ratio=0 \
--osd-scrub-interval-randomize-ratio=0"
for id in $(seq 0 2) ; do
run_osd $dir $id $ceph_osd_args || return 1
done
create_pool $poolname 1 1 || return 1
ceph osd pool set $poolname size 2
wait_for_clean || return 1
# Put an object
local payload=ABCDEF
echo $payload > $dir/ORIGINAL
rados --pool $poolname put SOMETHING $dir/ORIGINAL || return 1
# Remove the object from one shard physically
# Restarted osd get $ceph_osd_args passed
objectstore_tool $dir $(get_not_primary $poolname SOMETHING) SOMETHING remove || return 1
local pgid=$(get_pg $poolname SOMETHING)
local primary=$(get_primary $poolname SOMETHING)
local last_scrub_stamp="$(get_last_scrub_stamp $pgid)"
ceph tell $pgid deep_scrub
ceph tell $pgid scrub
# Wait for auto repair
wait_for_scrub $pgid "$last_scrub_stamp" || return 1
wait_for_clean || return 1
ceph pg dump pgs
# Verify - the file should be back
# Restarted osd get $ceph_osd_args passed
objectstore_tool $dir $(get_not_primary $poolname SOMETHING) SOMETHING list-attrs || return 1
objectstore_tool $dir $(get_not_primary $poolname SOMETHING) SOMETHING get-bytes $dir/COPY || return 1
diff $dir/ORIGINAL $dir/COPY || return 1
grep scrub_finish $dir/osd.${primary}.log
# Tear down
teardown $dir || return 1
}
function TEST_auto_repair_bluestore_scrub() {
local dir=$1
local poolname=testpool
# Launch a cluster with 5 seconds scrub interval
setup $dir || return 1
run_mon $dir a || return 1
run_mgr $dir x || return 1
local ceph_osd_args="--osd-scrub-auto-repair=true \
--osd_deep_scrub_randomize_ratio=0 \
--osd-scrub-interval-randomize-ratio=0"
for id in $(seq 0 2) ; do
run_osd $dir $id $ceph_osd_args || return 1
done
create_pool $poolname 1 1 || return 1
ceph osd pool set $poolname size 2
wait_for_clean || return 1
# Put an object
local payload=ABCDEF
echo $payload > $dir/ORIGINAL
rados --pool $poolname put SOMETHING $dir/ORIGINAL || return 1
# Remove the object from one shard physically
# Restarted osd get $ceph_osd_args passed
objectstore_tool $dir $(get_not_primary $poolname SOMETHING) SOMETHING remove || return 1
local pgid=$(get_pg $poolname SOMETHING)
local primary=$(get_primary $poolname SOMETHING)
local last_scrub_stamp="$(get_last_scrub_stamp $pgid)"
ceph tell $pgid scrub
# Wait for scrub -> auto repair
wait_for_scrub $pgid "$last_scrub_stamp" || return 1
ceph pg dump pgs
# Actually this causes 2 scrubs, so we better wait a little longer
sleep 5
wait_for_clean || return 1
ceph pg dump pgs
# Verify - the file should be back
# Restarted osd get $ceph_osd_args passed
objectstore_tool $dir $(get_not_primary $poolname SOMETHING) SOMETHING list-attrs || return 1
rados --pool $poolname get SOMETHING $dir/COPY || return 1
diff $dir/ORIGINAL $dir/COPY || return 1
grep scrub_finish $dir/osd.${primary}.log
# This should have caused 1 object to be repaired
COUNT=$(ceph pg $pgid query | jq '.info.stats.stat_sum.num_objects_repaired')
test "$COUNT" = "1" || return 1
# Tear down
teardown $dir || return 1
}
function TEST_auto_repair_bluestore_failed() {
local dir=$1
local poolname=testpool
# Launch a cluster with 5 seconds scrub interval
setup $dir || return 1
run_mon $dir a || return 1
run_mgr $dir x || return 1
local ceph_osd_args="--osd-scrub-auto-repair=true \
--osd_deep_scrub_randomize_ratio=0 \
--osd-scrub-interval-randomize-ratio=0"
for id in $(seq 0 2) ; do
run_osd $dir $id $ceph_osd_args || return 1
done
create_pool $poolname 1 1 || return 1
ceph osd pool set $poolname size 2
wait_for_clean || return 1
# Put an object
local payload=ABCDEF
echo $payload > $dir/ORIGINAL
for i in $(seq 1 10)
do
rados --pool $poolname put obj$i $dir/ORIGINAL || return 1
done
# Remove the object from one shard physically
# Restarted osd get $ceph_osd_args passed
objectstore_tool $dir $(get_not_primary $poolname SOMETHING) obj1 remove || return 1
# obj2 can't be repaired
objectstore_tool $dir $(get_not_primary $poolname SOMETHING) obj2 remove || return 1
objectstore_tool $dir $(get_primary $poolname SOMETHING) obj2 rm-attr _ || return 1
local pgid=$(get_pg $poolname obj1)
local primary=$(get_primary $poolname obj1)
local last_scrub_stamp="$(get_last_scrub_stamp $pgid)"
ceph tell $pgid deep_scrub
ceph tell $pgid scrub
# Wait for auto repair
wait_for_scrub $pgid "$last_scrub_stamp" || return 1
wait_for_clean || return 1
flush_pg_stats
grep scrub_finish $dir/osd.${primary}.log
grep -q "scrub_finish.*still present after re-scrub" $dir/osd.${primary}.log || return 1
ceph pg dump pgs
ceph pg dump pgs | grep -q "^${pgid}.*+failed_repair" || return 1
# Verify - obj1 should be back
# Restarted osd get $ceph_osd_args passed
objectstore_tool $dir $(get_not_primary $poolname obj1) obj1 list-attrs || return 1
rados --pool $poolname get obj1 $dir/COPY || return 1
diff $dir/ORIGINAL $dir/COPY || return 1
grep scrub_finish $dir/osd.${primary}.log
# Make it repairable
objectstore_tool $dir $(get_primary $poolname SOMETHING) obj2 remove || return 1
repair $pgid
sleep 2
flush_pg_stats
ceph pg dump pgs
ceph pg dump pgs | grep -q -e "^${pgid}.* active+clean " -e "^${pgid}.* active+clean+wait " || return 1
grep scrub_finish $dir/osd.${primary}.log
# Tear down
teardown $dir || return 1
}
function TEST_auto_repair_bluestore_failed_norecov() {
local dir=$1
local poolname=testpool
# Launch a cluster with 5 seconds scrub interval
setup $dir || return 1
run_mon $dir a || return 1
run_mgr $dir x || return 1
local ceph_osd_args="--osd-scrub-auto-repair=true \
--osd_deep_scrub_randomize_ratio=0 \
--osd-scrub-interval-randomize-ratio=0"
for id in $(seq 0 2) ; do
run_osd $dir $id $ceph_osd_args || return 1
done
create_pool $poolname 1 1 || return 1
ceph osd pool set $poolname size 2
wait_for_clean || return 1
# Put an object
local payload=ABCDEF
echo $payload > $dir/ORIGINAL
for i in $(seq 1 10)
do
rados --pool $poolname put obj$i $dir/ORIGINAL || return 1
done
# Remove the object from one shard physically
# Restarted osd get $ceph_osd_args passed
# obj1 can't be repaired
objectstore_tool $dir $(get_not_primary $poolname SOMETHING) obj1 remove || return 1
objectstore_tool $dir $(get_primary $poolname SOMETHING) obj1 rm-attr _ || return 1
# obj2 can't be repaired
objectstore_tool $dir $(get_not_primary $poolname SOMETHING) obj2 remove || return 1
objectstore_tool $dir $(get_primary $poolname SOMETHING) obj2 rm-attr _ || return 1
local pgid=$(get_pg $poolname obj1)
local primary=$(get_primary $poolname obj1)
local last_scrub_stamp="$(get_last_scrub_stamp $pgid)"
ceph tell $pgid deep_scrub
ceph tell $pgid scrub
# Wait for auto repair
wait_for_scrub $pgid "$last_scrub_stamp" || return 1
wait_for_clean || return 1
flush_pg_stats
grep -q "scrub_finish.*present with no repair possible" $dir/osd.${primary}.log || return 1
ceph pg dump pgs
ceph pg dump pgs | grep -q "^${pgid}.*+failed_repair" || return 1
# Tear down
teardown $dir || return 1
}
function TEST_repair_stats() {
local dir=$1
local poolname=testpool
local OSDS=2
local OBJS=30
# This need to be an even number
local REPAIRS=20
# Launch a cluster with 5 seconds scrub interval
setup $dir || return 1
run_mon $dir a || return 1
run_mgr $dir x || return 1
local ceph_osd_args="--osd_deep_scrub_randomize_ratio=0 \
--osd-scrub-interval-randomize-ratio=0"
for id in $(seq 0 $(expr $OSDS - 1)) ; do
run_osd $dir $id $ceph_osd_args || return 1
done
create_pool $poolname 1 1 || return 1
ceph osd pool set $poolname size 2
wait_for_clean || return 1
# Put an object
local payload=ABCDEF
echo $payload > $dir/ORIGINAL
for i in $(seq 1 $OBJS)
do
rados --pool $poolname put obj$i $dir/ORIGINAL || return 1
done
# Remove the object from one shard physically
# Restarted osd get $ceph_osd_args passed
local other=$(get_not_primary $poolname obj1)
local pgid=$(get_pg $poolname obj1)
local primary=$(get_primary $poolname obj1)
kill_daemons $dir TERM osd.$other >&2 < /dev/null || return 1
kill_daemons $dir TERM osd.$primary >&2 < /dev/null || return 1
for i in $(seq 1 $REPAIRS)
do
# Remove from both osd.0 and osd.1
OSD=$(expr $i % 2)
_objectstore_tool_nodown $dir $OSD obj$i remove || return 1
done
activate_osd $dir $primary $ceph_osd_args || return 1
activate_osd $dir $other $ceph_osd_args || return 1
wait_for_clean || return 1
repair $pgid
wait_for_clean || return 1
ceph pg dump pgs
flush_pg_stats
# This should have caused 1 object to be repaired
ceph pg $pgid query | jq '.info.stats.stat_sum'
COUNT=$(ceph pg $pgid query | jq '.info.stats.stat_sum.num_objects_repaired')
test "$COUNT" = "$REPAIRS" || return 1
ceph pg dump --format=json-pretty | jq ".pg_map.osd_stats[] | select(.osd == $primary )"
COUNT=$(ceph pg dump --format=json-pretty | jq ".pg_map.osd_stats[] | select(.osd == $primary ).num_shards_repaired")
test "$COUNT" = "$(expr $REPAIRS / 2)" || return 1
ceph pg dump --format=json-pretty | jq ".pg_map.osd_stats[] | select(.osd == $other )"
COUNT=$(ceph pg dump --format=json-pretty | jq ".pg_map.osd_stats[] | select(.osd == $other ).num_shards_repaired")
test "$COUNT" = "$(expr $REPAIRS / 2)" || return 1
ceph pg dump --format=json-pretty | jq ".pg_map.osd_stats_sum"
COUNT=$(ceph pg dump --format=json-pretty | jq ".pg_map.osd_stats_sum.num_shards_repaired")
test "$COUNT" = "$REPAIRS" || return 1
# Tear down
teardown $dir || return 1
}
function TEST_repair_stats_ec() {
local dir=$1
local poolname=testpool
local OSDS=3
local OBJS=30
# This need to be an even number
local REPAIRS=26
local allow_overwrites=false
# Launch a cluster with 5 seconds scrub interval
setup $dir || return 1
run_mon $dir a || return 1
run_mgr $dir x || return 1
local ceph_osd_args="--osd_deep_scrub_randomize_ratio=0 \
--osd-scrub-interval-randomize-ratio=0"
for id in $(seq 0 $(expr $OSDS - 1)) ; do
run_osd $dir $id $ceph_osd_args || return 1
done
# Create an EC pool
create_ec_pool $poolname $allow_overwrites k=2 m=1 || return 1
# Put an object
local payload=ABCDEF
echo $payload > $dir/ORIGINAL
for i in $(seq 1 $OBJS)
do
rados --pool $poolname put obj$i $dir/ORIGINAL || return 1
done
# Remove the object from one shard physically
# Restarted osd get $ceph_osd_args passed
local other=$(get_not_primary $poolname obj1)
local pgid=$(get_pg $poolname obj1)
local primary=$(get_primary $poolname obj1)
kill_daemons $dir TERM osd.$other >&2 < /dev/null || return 1
kill_daemons $dir TERM osd.$primary >&2 < /dev/null || return 1
for i in $(seq 1 $REPAIRS)
do
# Remove from both osd.0 and osd.1
OSD=$(expr $i % 2)
_objectstore_tool_nodown $dir $OSD obj$i remove || return 1
done
activate_osd $dir $primary $ceph_osd_args || return 1
activate_osd $dir $other $ceph_osd_args || return 1
wait_for_clean || return 1
repair $pgid
wait_for_clean || return 1
ceph pg dump pgs
flush_pg_stats
# This should have caused 1 object to be repaired
ceph pg $pgid query | jq '.info.stats.stat_sum'
COUNT=$(ceph pg $pgid query | jq '.info.stats.stat_sum.num_objects_repaired')
test "$COUNT" = "$REPAIRS" || return 1
for osd in $(seq 0 $(expr $OSDS - 1)) ; do
if [ $osd = $other -o $osd = $primary ]; then
repair=$(expr $REPAIRS / 2)
else
repair="0"
fi
ceph pg dump --format=json-pretty | jq ".pg_map.osd_stats[] | select(.osd == $osd )"
COUNT=$(ceph pg dump --format=json-pretty | jq ".pg_map.osd_stats[] | select(.osd == $osd ).num_shards_repaired")
test "$COUNT" = "$repair" || return 1
done
ceph pg dump --format=json-pretty | jq ".pg_map.osd_stats_sum"
COUNT=$(ceph pg dump --format=json-pretty | jq ".pg_map.osd_stats_sum.num_shards_repaired")
test "$COUNT" = "$REPAIRS" || return 1
# Tear down
teardown $dir || return 1
}
function corrupt_and_repair_jerasure() {
local dir=$1
local allow_overwrites=$2
local poolname=ecpool
setup $dir || return 1
run_mon $dir a || return 1
run_mgr $dir x || return 1
for id in $(seq 0 3) ; do
if [ "$allow_overwrites" = "true" ]; then
run_osd $dir $id || return 1
else
run_osd_filestore $dir $id || return 1
fi
done
create_rbd_pool || return 1
wait_for_clean || return 1
create_ec_pool $poolname $allow_overwrites k=2 m=2 || return 1
corrupt_and_repair_erasure_coded $dir $poolname || return 1
teardown $dir || return 1
}
function TEST_corrupt_and_repair_jerasure_appends() {
corrupt_and_repair_jerasure $1 false
}
function TEST_corrupt_and_repair_jerasure_overwrites() {
if [ "$use_ec_overwrite" = "true" ]; then
corrupt_and_repair_jerasure $1 true
fi
}
function corrupt_and_repair_lrc() {
local dir=$1
local allow_overwrites=$2
local poolname=ecpool
setup $dir || return 1
run_mon $dir a || return 1
run_mgr $dir x || return 1
for id in $(seq 0 9) ; do
if [ "$allow_overwrites" = "true" ]; then
run_osd $dir $id || return 1
else
run_osd_filestore $dir $id || return 1
fi
done
create_rbd_pool || return 1
wait_for_clean || return 1
create_ec_pool $poolname $allow_overwrites k=4 m=2 l=3 plugin=lrc || return 1
corrupt_and_repair_erasure_coded $dir $poolname || return 1
teardown $dir || return 1
}
function TEST_corrupt_and_repair_lrc_appends() {
corrupt_and_repair_lrc $1 false
}
function TEST_corrupt_and_repair_lrc_overwrites() {
if [ "$use_ec_overwrite" = "true" ]; then
corrupt_and_repair_lrc $1 true
fi
}
function unfound_erasure_coded() {
local dir=$1
local allow_overwrites=$2
local poolname=ecpool
local payload=ABCDEF
setup $dir || return 1
run_mon $dir a || return 1
run_mgr $dir x || return 1
for id in $(seq 0 3) ; do
if [ "$allow_overwrites" = "true" ]; then
run_osd $dir $id || return 1
else
run_osd_filestore $dir $id || return 1
fi
done
create_ec_pool $poolname $allow_overwrites k=2 m=2 || return 1
add_something $dir $poolname || return 1
local primary=$(get_primary $poolname SOMETHING)
local -a osds=($(get_osds $poolname SOMETHING | sed -e "s/$primary//"))
local not_primary_first=${osds[0]}
local not_primary_second=${osds[1]}
local not_primary_third=${osds[2]}
#
# 1) remove the corresponding file from the OSDs
#
pids=""
run_in_background pids objectstore_tool $dir $not_primary_first SOMETHING remove
run_in_background pids objectstore_tool $dir $not_primary_second SOMETHING remove
run_in_background pids objectstore_tool $dir $not_primary_third SOMETHING remove
wait_background pids
return_code=$?
if [ $return_code -ne 0 ]; then return $return_code; fi
#
# 2) repair the PG
#
local pg=$(get_pg $poolname SOMETHING)
repair $pg
#
# 3) check pg state
#
# it may take a bit to appear due to mon/mgr asynchrony
for f in `seq 1 60`; do
ceph -s | grep "1/1 objects unfound" && break
sleep 1
done
ceph -s|grep "4 up" || return 1
ceph -s|grep "4 in" || return 1
ceph -s|grep "1/1 objects unfound" || return 1
teardown $dir || return 1
}
function TEST_unfound_erasure_coded_appends() {
unfound_erasure_coded $1 false
}
function TEST_unfound_erasure_coded_overwrites() {
if [ "$use_ec_overwrite" = "true" ]; then
unfound_erasure_coded $1 true
fi
}
#
# list_missing for EC pool
#
function list_missing_erasure_coded() {
local dir=$1
local allow_overwrites=$2
local poolname=ecpool
setup $dir || return 1
run_mon $dir a || return 1
run_mgr $dir x || return 1
for id in $(seq 0 2) ; do
if [ "$allow_overwrites" = "true" ]; then
run_osd $dir $id || return 1
else
run_osd_filestore $dir $id || return 1
fi
done
create_rbd_pool || return 1
wait_for_clean || return 1
create_ec_pool $poolname $allow_overwrites k=2 m=1 || return 1
# Put an object and remove the two shards (including primary)
add_something $dir $poolname MOBJ0 || return 1
local -a osds0=($(get_osds $poolname MOBJ0))
# Put another object and remove two shards (excluding primary)
add_something $dir $poolname MOBJ1 || return 1
local -a osds1=($(get_osds $poolname MOBJ1))
# Stop all osd daemons
for id in $(seq 0 2) ; do
kill_daemons $dir TERM osd.$id >&2 < /dev/null || return 1
done
id=${osds0[0]}
ceph-objectstore-tool --data-path $dir/$id \
MOBJ0 remove || return 1
id=${osds0[1]}
ceph-objectstore-tool --data-path $dir/$id \
MOBJ0 remove || return 1
id=${osds1[1]}
ceph-objectstore-tool --data-path $dir/$id \
MOBJ1 remove || return 1
id=${osds1[2]}
ceph-objectstore-tool --data-path $dir/$id \
MOBJ1 remove || return 1
for id in $(seq 0 2) ; do
activate_osd $dir $id >&2 || return 1
done
create_rbd_pool || return 1
wait_for_clean || return 1
# Get get - both objects should in the same PG
local pg=$(get_pg $poolname MOBJ0)
# Repair the PG, which triggers the recovering,
# and should mark the object as unfound
repair $pg
for i in $(seq 0 120) ; do
[ $i -lt 60 ] || return 1
matches=$(ceph pg $pg list_unfound | egrep "MOBJ0|MOBJ1" | wc -l)
[ $matches -eq 2 ] && break
done
teardown $dir || return 1
}
function TEST_list_missing_erasure_coded_appends() {
list_missing_erasure_coded $1 false
}
function TEST_list_missing_erasure_coded_overwrites() {
if [ "$use_ec_overwrite" = "true" ]; then
list_missing_erasure_coded $1 true
fi
}
#
# Corrupt one copy of a replicated pool
#
function TEST_corrupt_scrub_replicated() {
local dir=$1
local poolname=csr_pool
local total_objs=19
setup $dir || return 1
run_mon $dir a --osd_pool_default_size=2 || return 1
run_mgr $dir x || return 1
run_osd $dir 0 || return 1
run_osd $dir 1 || return 1
create_rbd_pool || return 1
wait_for_clean || return 1
create_pool foo 1 || return 1
create_pool $poolname 1 1 || return 1
wait_for_clean || return 1
for i in $(seq 1 $total_objs) ; do
objname=ROBJ${i}
add_something $dir $poolname $objname || return 1
rados --pool $poolname setomapheader $objname hdr-$objname || return 1
rados --pool $poolname setomapval $objname key-$objname val-$objname || return 1
done
# Increase file 1 MB + 1KB
dd if=/dev/zero of=$dir/new.ROBJ19 bs=1024 count=1025
rados --pool $poolname put $objname $dir/new.ROBJ19 || return 1
rm -f $dir/new.ROBJ19
local pg=$(get_pg $poolname ROBJ0)
local primary=$(get_primary $poolname ROBJ0)
# Compute an old omap digest and save oi
CEPH_ARGS='' ceph daemon $(get_asok_path osd.0) \
config set osd_deep_scrub_update_digest_min_age 0
CEPH_ARGS='' ceph daemon $(get_asok_path osd.1) \
config set osd_deep_scrub_update_digest_min_age 0
pg_deep_scrub $pg
for i in $(seq 1 $total_objs) ; do
objname=ROBJ${i}
# Alternate corruption between osd.0 and osd.1
local osd=$(expr $i % 2)
case $i in
1)
# Size (deep scrub data_digest too)
local payload=UVWXYZZZ
echo $payload > $dir/CORRUPT
objectstore_tool $dir $osd $objname set-bytes $dir/CORRUPT || return 1
;;
2)
# digest (deep scrub only)
local payload=UVWXYZ
echo $payload > $dir/CORRUPT
objectstore_tool $dir $osd $objname set-bytes $dir/CORRUPT || return 1
;;
3)
# missing
objectstore_tool $dir $osd $objname remove || return 1
;;
4)
# Modify omap value (deep scrub only)