blob: 6937cf227d719040428390f9c914d74b5e6c05f1 (
plain)
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
|
#!/bin/bash
. $(dirname $0)/../../include.rc
. $(dirname $0)/../../volume.rc
cleanup;
START_TIMESTAMP=`date +%s`
function disconnected_brick_count {
local vol=$1
$CLI volume heal $vol info | \
egrep -i '(transport|Socket is not connected)' | wc -l
}
TEST glusterd
TEST pidof glusterd
TEST $CLI volume create $V0 replica 2 $H0:$B0/${V0}{0,1,2,3,4,5}
TEST $CLI volume set $V0 cluster.background-self-heal-count 0
TEST $CLI volume set $V0 cluster.eager-lock off
TEST $CLI volume start $V0
TEST glusterfs --volfile-id=/$V0 --volfile-server=$H0 $M0 --attribute-timeout=0 --entry-timeout=0
TEST kill_brick $V0 $H0 $B0/${V0}0
TEST kill_brick $V0 $H0 $B0/${V0}2
TEST kill_brick $V0 $H0 $B0/${V0}4
cd $M0
HEAL_FILES=0
for i in {1..10}
do
dd if=/dev/urandom of=f bs=1024k count=10
HEAL_FILES=$(($HEAL_FILES+1)) #+1 for data/metadata self-heal of 'f'
mkdir a; cd a;
#+3 for metadata self-heal of 'a' one per subvolume of DHT
HEAL_FILES=$(($HEAL_FILES+3))
done
#+3 represents entry sh on "/", one per subvolume of DHT?
HEAL_FILES=$(($HEAL_FILES + 3))
cd ~
EXPECT "$HEAL_FILES" afr_get_pending_heal_count $V0
#When bricks are down, it says Transport End point Not connected for them
EXPECT "3" disconnected_brick_count $V0
#Create some stale indices and verify that they are not counted in heal info
#TO create stale index create and delete files when one brick is down in
#replica pair.
for i in {11..20}; do echo abc > $M0/$i; done
HEAL_FILES=$(($HEAL_FILES + 10)) #count extra 10 files
EXPECT "$HEAL_FILES" afr_get_pending_heal_count $V0
#delete the files now, so that stale indices will remain.
for i in {11..20}; do rm -f $M0/$i; done
#After deleting files they should not appear in heal info
HEAL_FILES=$(($HEAL_FILES - 10))
EXPECT "$HEAL_FILES" afr_get_pending_heal_count $V0
TEST ! $CLI volume heal $V0
TEST $CLI volume set $V0 cluster.self-heal-daemon off
TEST ! $CLI volume heal $V0
TEST ! $CLI volume heal $V0 full
TEST $CLI volume start $V0 force
TEST $CLI volume set $V0 cluster.self-heal-daemon on
EXPECT_WITHIN 20 "Y" glustershd_up_status
EXPECT_WITHIN 20 "1" afr_child_up_status_in_shd $V0 0
EXPECT_WITHIN 20 "1" afr_child_up_status_in_shd $V0 2
EXPECT_WITHIN 20 "1" afr_child_up_status_in_shd $V0 4
TEST $CLI volume heal $V0
sleep 5 #Until the heal-statistics command implementation
#check that this heals the contents partially
TEST [ $HEAL_FILES -gt $(afr_get_pending_heal_count $V0) ]
TEST $CLI volume heal $V0 full
EXPECT_WITHIN 30 "0" afr_get_pending_heal_count $V0
#Test that ongoing IO is not considered as Pending heal
(dd if=/dev/zero of=$M0/file1 bs=1k 2>/dev/null 1>/dev/null)&
back_pid1=$!;
(dd if=/dev/zero of=$M0/file2 bs=1k 2>/dev/null 1>/dev/null)&
back_pid2=$!;
(dd if=/dev/zero of=$M0/file3 bs=1k 2>/dev/null 1>/dev/null)&
back_pid3=$!;
(dd if=/dev/zero of=$M0/file4 bs=1k 2>/dev/null 1>/dev/null)&
back_pid4=$!;
(dd if=/dev/zero of=$M0/file5 bs=1k 2>/dev/null 1>/dev/null)&
back_pid5=$!;
EXPECT 0 afr_get_pending_heal_count $V0
kill -SIGTERM $back_pid1;
kill -SIGTERM $back_pid2;
kill -SIGTERM $back_pid3;
kill -SIGTERM $back_pid4;
kill -SIGTERM $back_pid5;
wait >/dev/null 2>&1;
#Test that volume heal info reports files even when self-heal
#options are disabled
TEST touch $M0/f
TEST mkdir $M0/d
#DATA
TEST $CLI volume set $V0 cluster.data-self-heal off
EXPECT "off" volume_option $V0 cluster.data-self-heal
TEST kill_brick $V0 $H0 $B0/${V0}0
TEST kill_brick $V0 $H0 $B0/${V0}2
TEST kill_brick $V0 $H0 $B0/${V0}4
echo abc > $M0/f
EXPECT 1 afr_get_pending_heal_count $V0
TEST $CLI volume start $V0 force
EXPECT_WITHIN 20 "Y" glustershd_up_status
EXPECT_WITHIN 20 "1" afr_child_up_status_in_shd $V0 0
EXPECT_WITHIN 20 "1" afr_child_up_status_in_shd $V0 2
EXPECT_WITHIN 20 "1" afr_child_up_status_in_shd $V0 4
TEST $CLI volume heal $V0
EXPECT_WITHIN 30 "0" afr_get_pending_heal_count $V0
TEST $CLI volume set $V0 cluster.data-self-heal on
#METADATA
TEST $CLI volume set $V0 cluster.metadata-self-heal off
EXPECT "off" volume_option $V0 cluster.metadata-self-heal
TEST kill_brick $V0 $H0 $B0/${V0}0
TEST kill_brick $V0 $H0 $B0/${V0}2
TEST kill_brick $V0 $H0 $B0/${V0}4
TEST chmod 777 $M0/f
EXPECT 1 afr_get_pending_heal_count $V0
TEST $CLI volume start $V0 force
EXPECT_WITHIN 20 "Y" glustershd_up_status
EXPECT_WITHIN 20 "1" afr_child_up_status_in_shd $V0 0
EXPECT_WITHIN 20 "1" afr_child_up_status_in_shd $V0 2
EXPECT_WITHIN 20 "1" afr_child_up_status_in_shd $V0 4
TEST $CLI volume heal $V0
EXPECT_WITHIN 30 "0" afr_get_pending_heal_count $V0
TEST $CLI volume set $V0 cluster.metadata-self-heal on
#ENTRY
TEST $CLI volume set $V0 cluster.entry-self-heal off
EXPECT "off" volume_option $V0 cluster.entry-self-heal
TEST kill_brick $V0 $H0 $B0/${V0}0
TEST kill_brick $V0 $H0 $B0/${V0}2
TEST kill_brick $V0 $H0 $B0/${V0}4
TEST touch $M0/d/a
# 4 if mtime/ctime is modified for d in bricks without a
# 2 otherwise
PENDING=$( afr_get_pending_heal_count $V0 )
TEST test $PENDING -eq 2 -o $PENDING -eq 4
TEST $CLI volume start $V0 force
EXPECT_WITHIN 20 "Y" glustershd_up_status
EXPECT_WITHIN 20 "1" afr_child_up_status_in_shd $V0 0
EXPECT_WITHIN 20 "1" afr_child_up_status_in_shd $V0 2
EXPECT_WITHIN 20 "1" afr_child_up_status_in_shd $V0 4
TEST $CLI volume heal $V0
EXPECT_WITHIN 30 "0" afr_get_pending_heal_count $V0
TEST $CLI volume set $V0 cluster.entry-self-heal on
#Negative test cases
#Fail volume does not exist case
TEST ! $CLI volume heal fail info
#Fail volume stopped case
TEST $CLI volume stop $V0
TEST ! $CLI volume heal $V0 info
#Fail non-replicate volume info
TEST $CLI volume delete $V0
TEST $CLI volume create $V0 $H0:$B0/${V0}{6}
TEST $CLI volume start $V0
TEST ! $CLI volume heal $V0 info
# Check for non Linux systems that we did not mess with directory offsets
TEST ! log_newer $START_TIMESTAMP "offset reused from another DIR"
cleanup
|
