ALT Linux Bugzilla – #53455

Версия пакета: 4.6.0-alt1

Актуально для всех версий postgresql*-pgpool-II.

Не воспроизводится в р11 на версии 4.5.4-alt1 (например, с postgresql16-pgpool-II).

Стенд: 2 хоста

Шаги воспроизведения:

1. На обоих хостах установить postgresql:

# PG_VERSION=17
# apt-get update && apt-get install -y postgresql${PG_VERSION}-server
# /etc/init.d/postgresql initdb (для пользователя postgres задать пароль 1)

На первом (master) хосте:

# cp /var/lib/pgsql/data/postgresql.conf /var/lib/pgsql/data/postgresql.conf-orig && \
sed -i -e "s/#listen_addresses =.*/listen_addresses = '*'/" \
-e "s/#port = 5432/port = 5432/" \
-e "s/#wal_level =.*/wal_level = replica/" \
-e "s/#max_wal_senders.*/max_wal_senders = 2/" \
/var/lib/pgsql/data/postgresql.conf && \
diff -u --color /var/lib/pgsql/data/postgresql.conf-orig /var/lib/pgsql/data/postgresql.conf

На втором хосте:

# cp /var/lib/pgsql/data/postgresql.conf /var/lib/pgsql/data/postgresql.conf-orig && \
sed -i -e "s/#listen_addresses =.*/listen_addresses = '*'/" \
-e "s/#port = 5432/port = 5432/" \
-e "s/#hot_standby =.*/hot_standby = on/" /var/lib/pgsql/data/postgresql.conf && \
diff -u --color /var/lib/pgsql/data/postgresql.conf-orig /var/lib/pgsql/data/postgresql.conf

На обоих хостах:
 
# echo "host all all 0.0.0.0/0 scram-sha-256" >> /var/lib/pgsql/data/pg_hba.conf
# systemctl start postgresql; systemctl status postgresql

На master хосте установить пакет (обратить внимание на https://bugzilla.altlinux.org/53450):

# apt-get install postgresql${PG_VERSION}-pgpool-II

Настроить конфигурацию:

# SLAVE_HOST_IP=X.X.X.X
# MASTER_HOST_IP=$(hostname -i) && echo $MASTER_HOST_IP

# cp /etc/pgpool/pgpool.conf /etc/pgpool/pgpool.conf-orig && cat > /etc/pgpool/pgpool.conf <<EOF
listen_addresses = '*'
port = 9999
unix_socket_directories = '/var/run/pgpool'
pcp_listen_addresses = '*'
pcp_port = 9898
pcp_socket_dir = '/var/run/pgpool'
backend_hostname0 = '$MASTER_HOST_IP'
backend_port0 = 5432
backend_weight0 = 1
backend_data_directory0 = '/var/lib/pgsql/data'
backend_flag0 = 'ALLOW_TO_FAILOVER'
backend_hostname1 = '$SLAVE_HOST_IP'
backend_port1 = 5432
backend_weight1 = 1
backend_data_directory1 = '/var/lib/pgsql/data'
backend_flag1 = 'ALLOW_TO_FAILOVER'
enable_pool_hba = on
pool_passwd = '/etc/pgpool/pool_passwd'
load_balance_mode = on
replication_mode = on
num_init_children = 32
max_pool = 4
failover_when_quorum_exists = off
sr_check_user = 'postgres'
sr_check_password = '1'
EOF

# echo "host all all 0.0.0.0/0 scram-sha-256" >> /etc/pgpool/pool_hba.conf
# echo postgres:1 > /etc/pgpool/pool_passwd
# echo "postgres:$(pg_md5 1)" > /etc/pgpool/pcp.conf && cat /etc/pgpool/pcp.conf

Запусить pgpool:

# systemctl start pgpool; journalctl -fu pgpool

Получаемый результат:

pgpool[4867]: 2025-03-14 15:56:41.011: sr_check_worker pid 4867: LOG:  get_query_result failed: status: -2
pgpool[4867]: 2025-03-14 15:56:41.011: sr_check_worker pid 4867: CONTEXT:  while checking replication time lag
pgpool[4867]: 2025-03-14 15:56:51.038: sr_check_worker pid 4867: LOG:  get_query_result failed: status: -2
pgpool[4867]: 2025-03-14 15:56:51.038: sr_check_worker pid 4867: CONTEXT:  while checking replication time lag

Так же возникает ошибка при выполнении команд pcp, например:

# pcp_node_count -h $MASTER_HOST_IP -U postgres
FATAL:  authentication failed for user "postgres"
DETAIL:  username and/or password does not match