Thread: Hot Recovery not working
We have a network tunnel setup between 2 sites. I have PostgreSQL 9.4 installed on computers on both networks. I am attempted to setup a hot standby between the 2. From Site 1 I can establish a connection to Site 2 via the pgAdmin tool New Server Registration button. But I cannot establish a connection from Site2 to Site1 in the same manner. I get an error "Server doesn't listen.... could not connect to server: Connection refused (0x0000274D/10061) is the server running on host "x.x.x.x" and accepting TCP/IP connections on port 5432". I can ping each machine from the other. I've also run tracert and get a successful connection so I don't think it is a networking issue. I've also performed netstat on each to make sure servers is listening for TCP connection on port 5432 and they both are. I have identical entries in the pg_hba.conf and postgresql.conf files (except for IP address - each file lists the other servers IP) - set according to the following steps I performed- from an article I found online (https://www.openmakesoftware.com/insights/postgresql-hot-standby-for-windows/ ). 1. On Master – In pgAdmin run: CREATE USER replicator REPLICATION LOGIN ENCRYPTED PASSWORD 'mypassword'; 2. On Master – In pgAdmin, edit the postgresql.conf and set the following: wal_level = hot_standby max_wal_senders = 3 checkpoint_segments = 8 wal_keep_segments = 8 3. On Master – In pgAdmin, edit the pg_hba.conf and set the following where 5.6.7.8 is the ip address of the slave: host replication replicator 5.6.7.8/32 trust 4. On Master – restart the postgres windows service 5. On Slave – stop the postgres windows service 6. On Slave – open a command prompt and cd to your Postgres directory (c:\program files\Postgresql\9.4) and run: rd /s “c:\program files\Postgresql\9.4\data” 7. On Slave – open a command prompt and cd to your Postgres directory (c:\program files\Postgresql\9.4) and run where 1.2.3.4 is the ip of the Master: pg_basebackup -h 1.2.3.4 -D “c:\program files\Postgresql\9.4\data” -U replicator -v –P –X stream 8. On slave – In pgAdmin, edit the postgresql.conf and set the following: hot_standby = on 9. On slave – create a text file (c:\program files\Postgresql\9.4\data\recovery.conf) with the following contents where 1.2.3.4 is the ip of the Master: standby_mode = 'on'primary_conninfo = 'host=1.2.3.4 port=5432 user=replicator password=mypassword'trigger_file = 'c:\temp\postgresql.trigger' 10. On Slave – start the postgres windows service -- View this message in context: http://postgresql.nabble.com/Hot-Recovery-not-working-tp5892637.html Sent from the PostgreSQL - pgadmin support mailing list archive at Nabble.com.
You may want to check your firewall rules. A misconfigured switch/firewall could allow you to ping both ways just fine andcommunicate one way. -----Original Message----- From: pgadmin-support-owner@postgresql.org [mailto:pgadmin-support-owner@postgresql.org] On Behalf Of sreynolds Sent: Monday, March 14, 2016 7:23 AM To: pgadmin-support@postgresql.org Subject: [pgadmin-support] Hot Recovery not working We have a network tunnel setup between 2 sites. I have PostgreSQL 9.4 installed on computers on both networks. I am attemptedto setup a hot standby between the 2. From Site 1 I can establish a connection to Site 2 via the pgAdmin tool NewServer Registration button. But I cannot establish a connection from Site2 to Site1 in the same manner. I get an error"Server doesn't listen.... could not connect to server: Connection refused (0x0000274D/10061) is the server running on host "x.x.x.x" and accepting TCP/IP connections on port 5432". I can ping each machine from the other. I've also run tracert and get a successful connection so I don't think it is a networkingissue. I've also performed netstat on each to make sure servers is listening for TCP connection on port 5432 andthey both are. I have identical entries in the pg_hba.conf and postgresql.conf files (except for IP address - each file lists the otherservers IP) - set according to the following steps I performed- from an article I found online (https://www.openmakesoftware.com/insights/postgresql-hot-standby-for-windows/ ). 1. On Master – In pgAdmin run: CREATE USER replicator REPLICATION LOGIN ENCRYPTED PASSWORD 'mypassword'; 2. On Master – In pgAdmin, edit the postgresql.confand set the following: wal_level = hot_standby max_wal_senders = 3 checkpoint_segments = 8 wal_keep_segments = 8 3. On Master – In pgAdmin, edit the pg_hba.conf and set the following where 5.6.7.8 is the ip address of the slave: host replication replicator 5.6.7.8/32 trust 4. On Master – restart the postgres windows service 5. On Slave – stop the postgres windows service 6. On Slave – open acommand prompt and cd to your Postgres directory (c:\program files\Postgresql\9.4) and run: rd /s “c:\program files\Postgresql\9.4\data” 7. On Slave – open a command prompt and cd to your Postgres directory (c:\program files\Postgresql\9.4) and run where 1.2.3.4is the ip of the Master: pg_basebackup -h 1.2.3.4 -D “c:\program files\Postgresql\9.4\data” -U replicator -v –P –X stream 8. On slave – In pgAdmin,edit the postgresql.conf and set the following: hot_standby = on 9. On slave – create a text file (c:\program files\Postgresql\9.4\data\recovery.conf) with the following contents where 1.2.3.4 is the ip of the Master: standby_mode = 'on'primary_conninfo = 'host=1.2.3.4 port=5432 user=replicator password=mypassword'trigger_file = 'c:\temp\postgresql.trigger' 10. On Slave – start the postgres windows service -- View this message in context: http://postgresql.nabble.com/Hot-Recovery-not-working-tp5892637.html Sent from the PostgreSQL - pgadmin support mailing list archive at Nabble.com. -- Sent via pgadmin-support mailing list (pgadmin-support@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgadmin-support
I will talk to my network guy. We pass SQLServer data between the 2 so would think the rules are OK but I will verify. Thank you. ________________________________________ From: Anderson, Derek [Derek.Anderson@mspmac.org] Sent: Monday, March 14, 2016 12:46 PM To: Reynolds, Selwyn; pgadmin-support@postgresql.org Subject: RE: [pgadmin-support] Hot Recovery not working You may want to check your firewall rules. A misconfigured switch/firewall could allow you to ping both ways just fine andcommunicate one way. -----Original Message----- From: pgadmin-support-owner@postgresql.org [mailto:pgadmin-support-owner@postgresql.org] On Behalf Of sreynolds Sent: Monday, March 14, 2016 7:23 AM To: pgadmin-support@postgresql.org Subject: [pgadmin-support] Hot Recovery not working We have a network tunnel setup between 2 sites. I have PostgreSQL 9.4 installed on computers on both networks. I am attemptedto setup a hot standby between the 2. From Site 1 I can establish a connection to Site 2 via the pgAdmin tool NewServer Registration button. But I cannot establish a connection from Site2 to Site1 in the same manner. I get an error"Server doesn't listen.... could not connect to server: Connection refused (0x0000274D/10061) is the server running on host "x.x.x.x" and accepting TCP/IP connections on port 5432". I can ping each machine from the other. I've also run tracert and get a successful connection so I don't think it is a networkingissue. I've also performed netstat on each to make sure servers is listening for TCP connection on port 5432 andthey both are. I have identical entries in the pg_hba.conf and postgresql.conf files (except for IP address - each file lists the otherservers IP) - set according to the following steps I performed- from an article I found online (https://www.openmakesoftware.com/insights/postgresql-hot-standby-for-windows/ ). 1. On Master – In pgAdmin run: CREATE USER replicator REPLICATION LOGIN ENCRYPTED PASSWORD 'mypassword'; 2. On Master – In pgAdmin, edit the postgresql.confand set the following: wal_level = hot_standby max_wal_senders = 3 checkpoint_segments = 8 wal_keep_segments = 8 3. On Master – In pgAdmin, edit the pg_hba.conf and set the following where 5.6.7.8 is the ip address of the slave: host replication replicator 5.6.7.8/32 trust 4. On Master – restart the postgres windows service 5. On Slave – stop the postgres windows service 6. On Slave – open acommand prompt and cd to your Postgres directory (c:\program files\Postgresql\9.4) and run: rd /s “c:\program files\Postgresql\9.4\data” 7. On Slave – open a command prompt and cd to your Postgres directory (c:\program files\Postgresql\9.4) and run where 1.2.3.4is the ip of the Master: pg_basebackup -h 1.2.3.4 -D “c:\program files\Postgresql\9.4\data” -U replicator -v –P –X stream 8. On slave – In pgAdmin,edit the postgresql.conf and set the following: hot_standby = on 9. On slave – create a text file (c:\program files\Postgresql\9.4\data\recovery.conf) with the following contents where 1.2.3.4 is the ip of the Master: standby_mode = 'on'primary_conninfo = 'host=1.2.3.4 port=5432 user=replicator password=mypassword'trigger_file = 'c:\temp\postgresql.trigger' 10. On Slave – start the postgres windows service -- View this message in context: http://postgresql.nabble.com/Hot-Recovery-not-working-tp5892637.html Sent from the PostgreSQL - pgadmin support mailing list archive at Nabble.com. -- Sent via pgadmin-support mailing list (pgadmin-support@postgresql.org) To make changes to your subscription: http://www.postgresql.org/mailpref/pgadmin-support