Sorry, I have cancelled the previous message, so I am sending it once again.
Master DB server hangs during stopping in two nodes Suse Linux Enterprise Sever 12 (SLES12) cluster. The problem is reproducible on PostgreSQL 9.6.8 and PostgreSQL 10.4. PosgreSQL 10.4 faces with such a problem more frequently than 9.6.8.
PostgreSQL 10.4 was downloaded from https://www.postgresql.org/ftp/source/, configured and installed on SLES 12 system as follows: ./configure –prefix=usr Make Make install
psql --version psql (PostgreSQL) 10.4
Our configuration: I have configured 2-nodes database cluster( master (node01) and standby (node02)). The database on the master node was configured as follows:
Configuration files postgresql.conf and pg_hba.conf from the master and standby nodes have been attached to the mail. I was able to get some information about the replication by the following sql queries:
select * from pg_stat_replication # on master node select * from pg_stat_wal_receiver # on standby node
In order to stop database server on master node I invoked the command at 07:15:49:
pg_ctl –D /var/PostgresDb stop –m fast.
The most of the postgres processes were terminated but some of them remained to be working. Here is the list of postgres processes on the master node: