[Pgpool-general] pcp_attach_node problem?

Tatsuo Ishii ishii at sraoss.co.jp
Fri Jan 9 01:23:41 UTC 2009


Hum, stiil did not reproduce here. Could you show me a debug log of
pgpool?
--
Tatsuo Ishii
SRA OSS, Inc. Japan

> Well, I tested installing the latest CVS version in a clean box and this
> is what I get:
> 
> 1. [Server A] Check the status of the nodes:
> [root ~]# pcp_node_info 10 localhost 9898 postgres postgres 0
> 172.20.79.163 5432 2 1073741823.500000
> [root ~]# pcp_node_info 10 localhost 9898 postgres postgres 1
> 172.20.79.25 5432 2 1073741823.500000
> 
> 2. [Server B] stop second backend:
> service postgresql restart
> Stopping postgresql service:                               [  OK  ]
> Starting postgresql service:                               [  OK  ]
> 
> 3. [Server A] Check the status of the nodes:
> [root ~]# pcp_node_info 10 localhost 9898 postgres postgres 0
> 172.20.79.163 5432 2 1073741823.500000
> [root ~]# pcp_node_info 10 localhost 9898 postgres postgres 1
> 172.20.79.25 5432 3 1073741823.500000
> 
> 4. [Server A] Re-Attach node 1:
> pcp_attach_node 10 localhost 9898 postgres postgres 1
>  
> 5. [Server A] Check the status of the nodes:
> [root ~]# pcp_node_info 10 localhost 9898 postgres postgres 0
> 172.20.79.163 5432 2 1073741823.500000
> [root ~]# pcp_node_info 10 localhost 9898 postgres postgres 1
> 172.20.79.25 5432 1 1073741823.500000
> 
> 6. Here comes the problem (slightly different than before):
>  
>     If I had an already opened psql console, like:
>     [root ~]# psql -U postgres -p 9999
>     Welcome to psql 8.2.1, the PostgreSQL interactive terminal.
>  
>     Type:  \copyright for distribution terms
>         \h for help with SQL commands
>         \? for help with psql commands
>         \g or terminate with semicolon to execute query
>         \q to quit
>  
>     postgres=#
>  
>         And issue a SQL Select command on a table, like:
>             postgres=# select * from pg_stat_activity ;
>  
>     It returns:
>     postgres=# select 1;
>     server closed the connection unexpectedly
>         This probably means the server terminated abnormally
>         before or while processing the request.
>     The connection to the server was lost. Attempting reset: Succeeded.
> 
>     postgres=# select 1;
>     
>     Then it blocks. Meanwhile, when I try to check the statuses of both
> backends, pcp_* commands are also blocking.
> 
> Thanks,
> Daniel
> 
> -----Original Message-----
> From: pgpool-general-bounces at pgfoundry.org
> [mailto:pgpool-general-bounces at pgfoundry.org] On Behalf Of Tatsuo Ishii
> Sent: Wednesday, January 07, 2009 11:24 PM
> To: Crespo, Daniel @ SDS
> Cc: pgpool-general at pgfoundry.org
> Subject: Re: [Pgpool-general] pcp_attach_node problem?
> 
> Did not reproduce here?
> --
> Tatsuo Ishii
> SRA OSS, Inc. Japan
> 
> > Let me look into this...
> > --
> > Tatsuo Ishii
> > SRA OSS, Inc. Japan
> > 
> > > Finally, I installed the latest CVS checkout. I uninstalled my
> previous
> > > pgpool-II build by doing:
> > > 
> > > [root ~]# make uninstall
> > > [user ~]# make clean
> > > 
> > > Then, went to the just downloaded repository for pgpool-II and did:
> > > 
> > > [user ~]# ./configure
> > > [user ~]# make
> > > [root ~]# make install
> > > 
> > > Tried the below same thing, but now db connection through pgpool
> blocks
> > > as soon as second backend goes from status 2 to 3.
> > > 
> > > Anything wrong with the way I installed the most recent CVS head?
> > > 
> > > Thanks,
> > > 
> > > Daniel
> > > 
> > > 
> > > -----Original Message-----
> > > From: Tatsuo Ishii [mailto:ishii at sraoss.co.jp] 
> > > Sent: Wednesday, January 07, 2009 8:39 AM
> > > To: Crespo, Daniel @ SDS
> > > Cc: pgpool-general at pgfoundry.org
> > > Subject: Re: [Pgpool-general] pcp_attach_node problem?
> > > 
> > > Thanks for the report. I think it's a bug with CVS Head. pgpool
> should
> > > fail over in this case. Fix was committed. Could you please try it
> > > out?
> > > --
> > > Tatsuo Ishii
> > > SRA OSS, Inc. Japan
> > > 
> > > > Hello,
> > > >  
> > > > I have pgpool configured with two backends.
> > > >  
> > > > Below is the sequence of events to reproduce the
> > > seems-to-be-a-problem:
> > > >  
> > > > 1. [Server A] Check the status of the nodes:
> > > > [root ~]# pcp_node_info 10 localhost 9898 postgres postgres 0
> > > > 172.20.79.163 5432 2 1073741823.500000
> > > > [root ~]# pcp_node_info 10 localhost 9898 postgres postgres 1
> > > > 172.20.79.25 5432 2 1073741823.500000
> > > > 
> > > > 2. [Server B] stop second backend:
> > > > service postgresql restart
> > > > Stopping postgresql service:                               [  OK
> ]
> > > > Starting postgresql service:                               [  OK
> ]
> > > > 
> > > > 3. [Server A] Check the status of the nodes:
> > > > [root ~]# pcp_node_info 10 localhost 9898 postgres postgres 0
> > > > 172.20.79.163 5432 2 1073741823.500000
> > > > [root ~]# pcp_node_info 10 localhost 9898 postgres postgres 1
> > > > 172.20.79.25 5432 3 1073741823.500000
> > > > 
> > > > 4. [Server A] Re-Attach node 1:
> > > > pcp_attach_node 10 localhost 9898 postgres postgres 1
> > > >  
> > > > 5. [Server A] Check the status of the nodes:
> > > > [root ~]# pcp_node_info 10 localhost 9898 postgres postgres 0
> > > > 172.20.79.163 5432 2 1073741823.500000
> > > > [root ~]# pcp_node_info 10 localhost 9898 postgres postgres 1
> > > > 172.20.79.25 5432 1 1073741823.500000
> > > > 
> > > > 6. Here comes the problem:
> > > >  
> > > >     6.a If I had an already opened psql console, like:
> > > >         [root ~]# psql -U postgres -p 9999
> > > >         Welcome to psql 8.2.1, the PostgreSQL interactive
> terminal.
> > > >  
> > > >         Type:  \copyright for distribution terms
> > > >            \h for help with SQL commands
> > > >            \? for help with psql commands
> > > >            \g or terminate with semicolon to execute query
> > > >            \q to quit
> > > >  
> > > >         postgres=#
> > > >  
> > > >         And issue a SQL Select command on a table, like:
> > > >             postgres=# select * from pg_stat_activity ;
> > > >  
> > > >         It sits there forever. Meanwhile, the status of both
> backends
> > > > are still 2 and 1, respectively.
> > > >  
> > > >         If, instead, I disconnect the psql console, and connect it
> > > > again, both statuses become 2, and then I can issue the same SQL
> > > Select,
> > > > and works just fine.
> > > >  
> > > > Any clue?
> > > >  
> > > > Thanks,
> > > > Daniel
> > _______________________________________________
> > Pgpool-general mailing list
> > Pgpool-general at pgfoundry.org
> > http://pgfoundry.org/mailman/listinfo/pgpool-general
> _______________________________________________
> Pgpool-general mailing list
> Pgpool-general at pgfoundry.org
> http://pgfoundry.org/mailman/listinfo/pgpool-general
> _______________________________________________
> Pgpool-general mailing list
> Pgpool-general at pgfoundry.org
> http://pgfoundry.org/mailman/listinfo/pgpool-general


More information about the Pgpool-general mailing list