[Pgpool-general] online recovery fails on HPUX
Lazaro Rubén García Martinez
lgarciam at vnz.uci.cu
Wed Sep 14 13:41:08 UTC 2011
If you want md5 authentication between postgresql and pgpool, you need enable pool_hba property, configure pg_hba.conf and create a pool_passwd file. This file contains an user and password for connecting pgpool with database server. But, if md5 authentication is not used, for instance clear text password (password) the configuration can look like this:
In pgpool.conf
enable_pool_hba = false
# - online recovery -
# online recovery user
recovery_user = 'postgres'
# online recovery password
recovery_password = 'postgres'
In pg_hba.conf
# TYPE DATABASE USER CIDR-ADDRESS METHOD
host all all {Pgpool IP} password
In case of trust authentication the configuration looks like this:
In pgpool.conf
enable_pool_hba = false
# - online recovery -
# online recovery user
recovery_user = 'postgres'
# online recovery password
recovery_password = ''
In pg_hba.conf
# TYPE DATABASE USER CIDR-ADDRESS METHOD
host all all {Pgpool IP} trust
Regards
De: pgpool-general-bounces at pgfoundry.org [mailto:pgpool-general-bounces at pgfoundry.org] En nombre de Sandeep Thakkar
Enviado el: miércoles, 14 de septiembre de 2011 07:50
Para: Sandeep Thakkar; Jose Mendoza; pgpool-general at pgfoundry.org
Asunto: Re: [Pgpool-general] online recovery fails on HPUX
I have set "enable_pool_hba = false" in pgpool.conf and do not use pool_hba.conf. Do we need to enable this? How does the client authentication works when pool_hba.conf is disabled?
________________________________
From: Sandeep Thakkar <sandeeptt at yahoo.com>
To: Sandeep Thakkar <sandeeptt at yahoo.com>; Jose Mendoza <jose.mendoza at autonomy.com>; "pgpool-general at pgfoundry.org" <pgpool-general at pgfoundry.org>
Sent: Wednesday, September 14, 2011 4:33 PM
Subject: Re: [Pgpool-general] online recovery fails on HPUX
I still face this issue and I wonder why do I see the following error in the pgpool log:
2011-09-14 04:48:52 LOG: pid 10268: starting recovering node 1
2011-09-14 04:48:53 ERROR: pid 10268: start_recover: could not connect master node.
Please help! Thanks
________________________________
From: Sandeep Thakkar <sandeeptt at yahoo.com>
To: Sandeep Thakkar <sandeeptt at yahoo.com>; Jose Mendoza <jose.mendoza at autonomy.com>; "pgpool-general at pgfoundry.org" <pgpool-general at pgfoundry.org>
Sent: Tuesday, September 13, 2011 4:52 PM
Subject: Re: [Pgpool-general] online recovery fails on HPUX
I mean, pcp_remote_start contains "$PGCTL -w -D $DESTDIR start". i.e without SSH
Sandeep.
________________________________
From: Sandeep Thakkar <sandeeptt at yahoo.com>
To: Jose Mendoza <jose.mendoza at autonomy.com>; "pgpool-general at pgfoundry.org" <pgpool-general at pgfoundry.org>
Sent: Tuesday, September 13, 2011 4:24 PM
Subject: Re: [Pgpool-general] online recovery fails on HPUX
Well, actually my database servers and pgpool running on the same host, so my pcp_remote_start does not contains "$PGCTL -w -D $DESTDIR start". i.e without SSH. This works fine on Linux, though.
Sandeep.
________________________________
From: Jose Mendoza <jose.mendoza at autonomy.com>
To: pgpool-general at pgfoundry.org
Sent: Friday, September 2, 2011 12:47 PM
Subject: Re: [Pgpool-general] online recovery fails on HPUX
If you already have an sshkey defined and it’s the same user on both servers then. I am not sure what could be causing the failure but according to your log its an ssh auth issue.
Have you tried running a telnet test to the port to check that is listening…
Jose
Autonomy Ops
verificare tua hinc
From: Sandeep Thakkar [mailto:sandeeptt at yahoo.com]
Sent: Friday, September 02, 2011 12:05 AM
To: Jose Mendoza; pgpool-general at pgfoundry.org
Subject: Re: [Pgpool-general] online recovery fails on HPUX
Sorry, I didn't get you.. Actually, both the server instances and the pgpool running on the same host.. Could this be an issue of loopback?
________________________________
From: Jose Mendoza <jose.mendoza at autonomy.com>
To: pgpool-general at pgfoundry.org
Sent: Friday, September 2, 2011 12:16 PM
Subject: Re: [Pgpool-general] online recovery fails on HPUX
SSh trust must be created on both server for all the accounts involved in the recovery process.
I would add the ssh-keys the the user for pgpool and try again.
Jose
Autonomy Ops
verificare tua hinc
From: Sandeep Thakkar [mailto:sandeeptt at yahoo.com]
Sent: Thursday, September 01, 2011 10:17 PM
To: Jose Mendoza; pgpool-general at pgfoundry.org
Subject: Re: [Pgpool-general] online recovery fails on HPUX
I can see the following lines in /tmp/pgpool.log:
......
2011-09-01 22:57:34 ERROR: pid 22975: check_replication_time_lag: DB node is valid but no persistent connection
2011-09-01 22:57:34 DEBUG: pid 22936: health_check: 1 th DB node status: 3
2011-09-01 22:57:36 LOG: pid 23019: starting recovering node 1
2011-09-01 22:57:36 ERROR: pid 23019: start_recover: could not connect master node.
2011-09-01 22:57:38 DEBUG: pid 22936: starting health checking
2011-09-01 22:57:38 DEBUG: pid 22936: health_check: 0 th DB node status: 1
2011-09-01 22:57:38 DEBUG: pid 22975: pool_ssl: SSL requested but SSL support is not available
2011-09-01 22:57:38 DEBUG: pid 22975: s_do_auth: auth kind: 0
2011-09-01 22:57:38 DEBUG: pid 22975: s_do_auth: parameter status data received
2011-09-01 22:57:38 DEBUG: pid 22975: s_do_auth: parameter status data received
2011-09-01 22:57:38 DEBUG: pid 22975: s_do_auth: backend key data received
2011-09-01 22:57:38 DEBUG: pid 22975: s_do_auth: transaction state: I
2011-09-01 22:57:39 DEBUG: pid 22936: health_check: 1 th DB node status: 3
2011-09-01 22:57:39 ERROR: pid 22975: connect_inet_domain_socket: connect() failed: Connection refused
2011-09-01 22:57:39 ERROR: pid 22975: make_persistent_db_connection: connection to localhost(5445) failed
2011-09-01 22:57:39 DEBUG: pid 22975: do_query: kind: T
2011-09-01 22:57:39 DEBUG: pid 22975: num_fileds: 1
2011-09-01 22:57:39 DEBUG: pid 22975: do_query: kind: D
2011-09-01 22:57:39 DEBUG: pid 22975: do_query: kind: C
2011-09-01 22:57:39 DEBUG: pid 22975: do_query: kind: Z
2011-09-01 22:57:39 ERROR: pid 22975: check_replication_time_lag: DB node is valid but no persistent connection
2011-09-01 22:57:43 DEBUG: pid 22936: starting health checking
......
________________________________
From: Sandeep Thakkar <sandeeptt at yahoo.com>
To: Jose Mendoza <jose.mendoza at autonomy.com>; "pgpool-general at pgfoundry.org" <pgpool-general at pgfoundry.org>
Sent: Friday, September 2, 2011 10:10 AM
Subject: Re: [Pgpool-general] online recovery fails on HPUX
# Logging directory
logdir = '/tmp'
________________________________
From: Jose Mendoza <jose.mendoza at autonomy.com>
To: pgpool-general at pgfoundry.org
Sent: Wednesday, August 31, 2011 10:28 PM
Subject: Re: [Pgpool-general] online recovery fails on HPUX
What does your pgpool.conf say about logging:
# Logging directory
logdir = '/var/log'
Jose
Autonomy Ops
verificare tua hinc
From: pgpool-general-bounces at pgfoundry.org [mailto:pgpool-general-bounces at pgfoundry.org] On Behalf Of Sandeep Thakkar
Sent: Wednesday, August 31, 2011 4:58 AM
To: Sandeep Thakkar; pgpool-general at pgfoundry.org
Subject: Re: [Pgpool-general] online recovery fails on HPUX
I see following two lines when the pcp_recovery_node is fired.. Also, I cannot find recovery.log, but recovery.conf is present in standby cluster directory..
....
2011-08-31 05:49:22 LOG: pid 20430: starting recovering node 1
2011-08-31 05:49:22 ERROR: pid 20430: start_recover: could not connect master node.
....
The pg_hba.conf has the following entry:
# "local" is for Unix domain socket connections only
local all all trust
# IPv4 local connections:
host all all 127.0.0.1/32 trust
# IPv6 local connections:
host all all ::1/128 trust
host replication edb 127.0.0.1/32 trust
________________________________
From: Sandeep Thakkar <sandeeptt at yahoo.com>
To: "pgpool-general at pgfoundry.org" <pgpool-general at pgfoundry.org>
Sent: Tuesday, August 30, 2011 2:44 PM
Subject: [Pgpool-general] online recovery fails on HPUX
Hi,
I'm using Postgres Plus Advanced Server on HP-UX 11.31 and found that the online recovery does not work here.
pgPool version is 3.0.3-4
$pcp_recovery_node -d 20 localhost 9898 edb edb 1
DEBUG: send: tos="R", len=41
DEBUG: recv: tos="r", len=21, data=AuthenticationOK
DEBUG: send: tos="D", len=6
DEBUG: recv: tos="e", len=20, data=recovery failed
DEBUG: command failed. reason=recovery failed
BackendError
DEBUG: send: tos="X", len=4
I can see the following in the log:
....
LOG: incomplete startup packet
LOG: connection failed during start up processing: user= database=
....
Moreover, I do not see pgpool_recovery_node statement in the log, so wondering if it is really getting executed. I have tried my scripts on Linux and it worked all fine. So, there is no issue with the pgpool.conf or my basebackup and other files.. Could this be an issue of HPUX?
Thanks.
_______________________________________________
Pgpool-general mailing list
Pgpool-general at pgfoundry.org<mailto:Pgpool-general at pgfoundry.org>
http://pgfoundry.org/mailman/listinfo/pgpool-general
_______________________________________________
Pgpool-general mailing list
Pgpool-general at pgfoundry.org<mailto:Pgpool-general at pgfoundry.org>
http://pgfoundry.org/mailman/listinfo/pgpool-general
_______________________________________________
Pgpool-general mailing list
Pgpool-general at pgfoundry.org<mailto:Pgpool-general at pgfoundry.org>
http://pgfoundry.org/mailman/listinfo/pgpool-general
_______________________________________________
Pgpool-general mailing list
Pgpool-general at pgfoundry.org<mailto:Pgpool-general at pgfoundry.org>
http://pgfoundry.org/mailman/listinfo/pgpool-general
_______________________________________________
Pgpool-general mailing list
Pgpool-general at pgfoundry.org<mailto:Pgpool-general at pgfoundry.org>
http://pgfoundry.org/mailman/listinfo/pgpool-general
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://pgfoundry.org/pipermail/pgpool-general/attachments/20110914/c8d467f2/attachment-0001.html>
More information about the Pgpool-general
mailing list