blob: 01878d7e589eee1564f02b70a972709c560fe741 (
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
|
#!/bin/bash
# This script is run by failover_command.
set -o xtrace
# Special values:
# 1) %d = failed node id
# 2) %h = failed node hostname
# 3) %p = failed node port number
# 4) %D = failed node database cluster path
# 5) %m = new main node id
# 6) %H = new main node hostname
# 7) %M = old main node id
# 8) %P = old primary node id
# 9) %r = new main port number
# 10) %R = new main database cluster path
# 11) %N = old primary node hostname
# 12) %S = old primary node port number
# 13) %% = '%' character
FAILED_NODE_ID="$1"
FAILED_NODE_HOST="$2"
FAILED_NODE_PORT="$3"
FAILED_NODE_PGDATA="$4"
NEW_MAIN_NODE_ID="$5"
NEW_MAIN_NODE_HOST="$6"
OLD_MAIN_NODE_ID="$7"
OLD_PRIMARY_NODE_ID="$8"
NEW_MAIN_NODE_PORT="$9"
NEW_MAIN_NODE_PGDATA="${10}"
OLD_PRIMARY_NODE_HOST="${11}"
OLD_PRIMARY_NODE_PORT="${12}"
PGHOME=/usr/pgsql-17
REPL_SLOT_NAME=$(echo ${FAILED_NODE_HOST,,} | tr -- -. _)
POSTGRESQL_STARTUP_USER=postgres
SSH_KEY_FILE=id_rsa_pgpool
SSH_OPTIONS="-o StrictHostKeyChecking=no -o UserKnownHostsFile=/dev/null -i ~/.ssh/${SSH_KEY_FILE}"
echo failover.sh: start: failed_node_id=$FAILED_NODE_ID failed_host=$FAILED_NODE_HOST \
old_primary_node_id=$OLD_PRIMARY_NODE_ID new_main_node_id=$NEW_MAIN_NODE_ID new_main_host=$NEW_MAIN_NODE_HOST
## If there's no main node anymore, skip failover.
if [ $NEW_MAIN_NODE_ID -lt 0 ]; then
echo failover.sh: All nodes are down. Skipping failover.
exit 0
fi
## Test passwordless SSH
ssh -T ${SSH_OPTIONS} ${POSTGRESQL_STARTUP_USER}@${NEW_MAIN_NODE_HOST} ls /tmp > /dev/null
if [ $? -ne 0 ]; then
echo failover.sh: passwordless SSH to ${POSTGRESQL_STARTUP_USER}@${NEW_MAIN_NODE_HOST} failed. Please setup passwordless SSH.
exit 1
fi
## If Standby node is down, skip failover.
if [ $OLD_PRIMARY_NODE_ID != "-1" -a $FAILED_NODE_ID != $OLD_PRIMARY_NODE_ID ]; then
# If Standby node is down, drop replication slot.
${PGHOME}/bin/psql -h ${OLD_PRIMARY_NODE_HOST} -p ${OLD_PRIMARY_NODE_PORT} postgres \
-c "SELECT pg_drop_replication_slot('${REPL_SLOT_NAME}');" >/dev/null 2>&1
if [ $? -ne 0 ]; then
echo ERROR: failover.sh: drop replication slot \"${REPL_SLOT_NAME}\" failed. You may need to drop replication slot manually.
fi
echo failover.sh: end: standby node is down. Skipping failover.
exit 0
fi
## Promote Standby node.
echo failover.sh: primary node is down, promote new_main_node_id=$NEW_MAIN_NODE_ID on ${NEW_MAIN_NODE_HOST}.
ssh -T ${SSH_OPTIONS} ${POSTGRESQL_STARTUP_USER}@${NEW_MAIN_NODE_HOST} ${PGHOME}/bin/pg_ctl -D ${NEW_MAIN_NODE_PGDATA} -w promote
if [ $? -ne 0 ]; then
echo ERROR: failover.sh: promote failed
exit 1
fi
echo failover.sh: end: new_main_node_id=$NEW_MAIN_NODE_ID on ${NEW_MAIN_NODE_HOST} was successfully promoted to primary
exit 0
|
