[Pgpool-general] start_recovery: remote start failed

[Daniel.Crespo at l-3com.com]

> Hi!
> 
> i have this trouble with pgpool-II 2.2.6  and  start_recovery.
> 
> 
> when i tried to recover  a node whit this command:
> 
> pcp_recovery_node 5 pgsql1 9898 root rutilio 1
> 
> and the log of pgpool is:
> 
> 2010-01-27 09:24:12 LOG:   pid 6036: CHECKPOINT in the 1st stage done
> 2010-01-27 09:24:12 LOG:   pid 6036: 1st stage is done
> 2010-01-27 09:24:12 LOG:   pid 6036: starting 2nd stage
> 2010-01-27 09:24:12 LOG:   pid 6036: all connections from clients have
> been closed
> 2010-01-27 09:24:12 LOG:   pid 6036: CHECKPOINT in the 2nd stage done
> 2010-01-27 09:24:12 ERROR: pid 6036: exec_remote_start:
> pgpool_remote_start failed: ERROR:  function pgpool_remote_start(u
> nknown, unknown) does not exist
> LINE 1: SELECT pgpool_remote_start('pgsql2', '')
> HINT:  No function matches the given name and argument types. You
might
> need to add explicit type casts.
> 
> 2010-01-27 09:24:12 ERROR: pid 6036: start_recovery: remote start
> failed
> 
> 
> What can  i do to solve this?  maybe i have missing  some settings???

Hi,

You need to install pgpool recovery stored procedures in all database
backends.
In your pgpool source dir, go to:
/usr/local/src/pgpool-II-2.2.6/sql/pgpool-recovery
There should be a file called pgpool-recovery.sql. Execute it in the
following way on all database backends:

psql -U postgres -d template1 -f pgpool-recovery.sql

If you already created a database, then execute it in that database by
replacing "template1" with the name of the database.

And try again.

Daniel