[Pgcluster-general] Recovery start

a.mitani at sra-europe.com a.mitani at sra-europe.com
Thu Jan 11 14:28:16 UTC 2007 

Hi Tom,

The purpose of "-U" option is very similar with "-R" option.
However calling command is different.
<rsync> is used by "-R" option, but <pg_dump & pg_restore> are used by
"-U" option.

In your case, it seems a problem of rsync configuration in Node A.

Regards,
------------------
At.Mitani

> It worked, the option was "-U" (uppercase). Though, if I use PGCluster
> in real world, can I avoid using -R (i.e. replacing it with -U) and be
> sure I'm not missing anything?
>
> Regards,
> --
> Tom;
>
> a.mitani at sra-europe.com escribi&#65533;:
>> It seems the Cluster A is failed to create backup at the recovery
>> preparation.
>>
>> Could you try recovery with "-u" option.
>> It's a cold recovery option without backup files.
>>
>> Regards,
>> -----------------
>> At.Mitani
>>
>>
>>> Wed Jan 10 13:16:48 2007  pgrecovery_loop():recovery accept port 8101
>>> Wed Jan 10 13:16:48 2007  read_packet():receive packet
>>> Wed Jan 10 13:16:48 2007  no = 1
>>> Wed Jan 10 13:16:48 2007  max_connect = 100
>>> Wed Jan 10 13:16:48 2007  port = 5432
>>> Wed Jan 10 13:16:48 2007  recoveryPort = 7001
>>> Wed Jan 10 13:16:48 2007  hostName = ldapserver
>>> Wed Jan 10 13:16:48 2007  pg_data = /usr/local/pgsql/data
>>> Wed Jan 10 13:16:48 2007  pgrecovery_loop():receive packet no:1
>>> Wed Jan 10 13:16:48 2007  pgrecovery_loop():1st master  - 0
>>> Wed Jan 10 13:16:48 2007  pgrecovery_loop():1st target  - 0
>>> Wed Jan 10 13:16:48 2007  first_setup_recovery():1st setup target
>>> ldapserver
>>> Wed Jan 10 13:16:48 2007  first_setup_recovery():1st setup port 5432
>>> Wed Jan 10 13:16:48 2007  first_setup_recovery():add recovery target to
>>> host table
>>> Wed Jan 10 13:16:48 2007  first_setup_recovery():set
>>> RECOVERY_PGDATA_REQ
>>> packet data
>>> Wed Jan 10 13:16:48 2007  pgr_createConn():PQsetdbLogin
>>> host[10.11.80.80] port[5432] db[template1] user[postgres]
>>> Wed Jan 10 13:16:48 2007  pgr_createConn():PQsetdbLogin ok!!
>>> Wed Jan 10 13:16:49 2007  PGRsend_replicate_packet_to_server():connect
>>> db:template1 port:5432 user:postgres host:10.11.80.80 query:VACUUM
>>> Wed Jan 10 13:16:49 2007
>>> send_replicate_packet_to_server():sync_command(SELECT
>>> PGR_SYSTEM_COMMAND_FUNCTION(3,0,0,0,1,1) )
>>> Wed Jan 10 13:16:49 2007
>>> send_replicate_packet_to_server():sync_command(SELECT
>>> PGR_SYSTEM_COMMAND_FUNCTION(8,0,0,1) )
>>> Wed Jan 10 13:16:49 2007
>>> send_replicate_packet_to_server():sync_command
>>> returns SYSTEM_COMMAND
>>> Wed Jan 10 13:16:49 2007  send_replicate_packet_to_server():PQexec send
>>> :VACUUM
>>> Wed Jan 10 13:16:49 2007  send_replicate_packet_to_server():PQexec
>>> returns :VACUUM
>>> Wed Jan 10 13:16:49 2007  first_setup_recovery():send packet to master
>>> ldapserver2.mecon.ar recoveryPort 7001
>>> Wed Jan 10 13:16:49 2007  first_setup_recovery():wait answer from
>>> master
>>> server
>>> Wed Jan 10 13:16:49 2007  read_packet():receive packet
>>> Wed Jan 10 13:16:49 2007  no = 3
>>> Wed Jan 10 13:16:49 2007  max_connect = 100
>>> Wed Jan 10 13:16:49 2007  port = 5432
>>> Wed Jan 10 13:16:49 2007  recoveryPort = 7001
>>> Wed Jan 10 13:16:49 2007  hostName = ldapserver2.mecon.ar
>>> Wed Jan 10 13:16:49 2007  pg_data = /usr/local/pgsql/data
>>> Wed Jan 10 13:16:49 2007  first_setup_recovery():get answer from
>>> master:no[3]
>>> Wed Jan 10 13:16:49 2007  pgrecovery_loop():first_setup_recovery end:0
>>> Wed Jan 10 13:16:49 2007  read_packet():receive packet
>>> Wed Jan 10 13:16:49 2007  replicate_loop():session closed
>>> Wed Jan 10 13:16:49 2007  no = 200
>>> Wed Jan 10 13:16:49 2007  max_connect = 100
>>> Wed Jan 10 13:16:49 2007  replicate_loop():replicate loop exit
>>> Wed Jan 10 13:16:49 2007  port = 5432
>>> Wed Jan 10 13:16:49 2007  recoveryPort = 7001
>>> Wed Jan 10 13:16:49 2007  hostName = ldapserver
>>> Wed Jan 10 13:16:49 2007  pg_data = /usr/local/pgsql/data
>>> Wed Jan 10 13:16:49 2007  pgrecovery_loop():receive packet no:200
>>> Wed Jan 10 13:16:49 2007  pgrecovery_loop():recovery error accept. top
>>> queueing and initiarse recovery status
>>> Wed Jan 10 13:16:49 2007  PGRsend_queue():master ldapserver2.mecon.ar -
>>> 5432
>>> Wed Jan 10 13:16:49 2007  PGRsend_queue():send_queue return status 0
>>>
>>>
>>> a.mitani at sra-europe.com escribi&#65533;:
>>>
>>>> I see...
>>>> Then, could you show us the debug message of replicator B during
>>>> recovery
>>>> of Cluster A.
>>>>
>>>> Regards,
>>>> -------------
>>>> At.Mitani
>>>>
>>>>
>>>>
>>>>> If you mean for me to execute some query that updates the DB, I've
>>>>> done
>>>>> it but it's all the same :(
>>>>>
>>>>> a.mitani at sra-europe.com escribi&#65533;:
>>>>>
>>>>>
>>>>>> Hi,
>>>>>>
>>>>>> Please try again after execute update query in Node B.
>>>>>>
>>>>>>
>>>>>>
>>>>>>
>>>>>>> I have the following error when trying to start a cluster in
>>>>>>> recovery
>>>>>>> mode:
>>>>>>>
>>>>>>> [postgres at ldapserver data]$ postmaster -R -D /usr/local/pgsql/data
>>>>>>> Start in recovery mode!
>>>>>>> Please wait until a data synchronization finishes from Master DB...
>>>>>>> 1st recovery step of [global] directory...rsync: link_stat
>>>>>>> "/usr/local/pgsql/data/global" failed: No such file or directory
>>>>>>> (2)
>>>>>>> rsync error: some files could not be transferred (code 23) at
>>>>>>> main.c(702)
>>>>>>> NG
>>>>>>> PGR_Recovery_Main() failed with cold recovery
>>>>>>>
>>>>>>> What could it be?
>>>>>>>
>>>>>>> Regards,
>>>>>>> --
>>>>>>> Tom;
>>>>>>> _______________________________________________
>>>>>>> Pgcluster-general mailing list
>>>>>>> Pgcluster-general at pgfoundry.org
>>>>>>> http://pgfoundry.org/mailman/listinfo/pgcluster-general
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>> _______________________________________________
>>>>>> Pgcluster-general mailing list
>>>>>> Pgcluster-general at pgfoundry.org
>>>>>> http://pgfoundry.org/mailman/listinfo/pgcluster-general
>>>>>>
>>>>>>
>>>>>>
>>>>> _______________________________________________
>>>>> Pgcluster-general mailing list
>>>>> Pgcluster-general at pgfoundry.org
>>>>> http://pgfoundry.org/mailman/listinfo/pgcluster-general
>>>>>
>>>>>
>>>>>
>>>> _______________________________________________
>>>> Pgcluster-general mailing list
>>>> Pgcluster-general at pgfoundry.org
>>>> http://pgfoundry.org/mailman/listinfo/pgcluster-general
>>>>
>>>>
>>> _______________________________________________
>>> Pgcluster-general mailing list
>>> Pgcluster-general at pgfoundry.org
>>> http://pgfoundry.org/mailman/listinfo/pgcluster-general
>>>
>>>
>>
>> _______________________________________________
>> Pgcluster-general mailing list
>> Pgcluster-general at pgfoundry.org
>> http://pgfoundry.org/mailman/listinfo/pgcluster-general
>>
>
> _______________________________________________
> Pgcluster-general mailing list
> Pgcluster-general at pgfoundry.org
> http://pgfoundry.org/mailman/listinfo/pgcluster-general
>

More information about the Pgcluster-general mailing list