[pgpool-general: 1727] Re: [PgPool-II 3.2.4] pgpool_regclass now mandatory?

Tatsuo Ishii ishii at postgresql.org
Tue May 14 08:09:13 JST 2013


>> Hello all.
>> 
>> I tried Pgpool-II 3.2.4 and it seems that the pgpool_regclass is now
>> mandatory?
>> Indeed pgpool have degenerated a node just after I put the 3.2.4 version in
>> production; here is the logs:
>> 
>> May 13 09:55:00 node1 pgpool-II[10714]: connection received: host=192.0.0.7
>> port=36766
>> May 13 09:55:00 node1 pgpool-II[10714]: do_query: error message from
>> backend function "pgpool_regclass(cstring)" does not exist
>> May 13 09:55:00 node1 pgpool-II[10714]: do_query: error message from
>> backend current transaction is aborted, commands ignored until end of
>> transaction block
>> May 13 09:55:00 node1 pgpool-II[10714]: do_query: error message from
>> backend current transaction is aborted, commands ignored until end of
>> transaction block
>> May 13 09:55:00 node1 pgpool-II[10714]: pool_send_and_wait: Error or notice
>> message from backend: : DB node id: 0 backend pid: 10736 statement: INSERT
>> INTO blablabla; message: current transaction is aborted, commands ignored
>> until end of transaction block
>> May 13 09:55:00 node1 pgpool-II[10714]: read_kind_from_backend: 1 th kind C
>> does not match with master or majority connection kind E
>> May 13 09:55:00 node1 pgpool-II[10714]: kind mismatch among backends.
>> Possible last query was: "INSERT INTO blablabla;" kind details are: 0[E:
>> current transaction is aborted, commands ignored until end of transaction
>> block] 1[C]
>> May 13 09:55:00 node1 pgpool-II[10714]: degenerate_backend_set: 1 fail over
>> request from pid 10714
>> May 13 09:55:00 node1 pgpool-II[10169]: starting degeneration. shutdown
>> host 192.0.0.19(5432)
>> May 13 09:55:00 node1 pgpool-II[10169]: Restart all children
>> May 13 09:55:00 node1 pgpool-II[10169]: execute command: sudo
>> /usr/bin/block_failed_db 1 &
>> May 13 09:55:00 node1 pgpool-II[10169]: failover: set new primary node: -1
>> May 13 09:55:00 node1 pgpool-II[10169]: failover: set new master node: 0
>> May 13 09:55:00 node1 pgpool-II[10169]: failover done. shutdown host
>> 192.0.0.19(5432)
>> May 13 09:55:00 node1 pgpool-II[10316]: worker process received restart
>> request
>> May 13 09:55:00 node1 pgpool-II[10756]: connection received:
>> host=192.0.0.117 port=54452
>> 
>> I never added the table to my databases before (even I know that it's
>> "strongly recommended" in the documentation) and I'm using PgPool since
>> several years now.
>> According to releases notes I guess this is related to the changes about
>> pool_has_pgpool_regclass.
>> 
>> So my question is: does I really need to add the pgpool_regclass in my
>> databases or it's a side effect of the 3.4.8 which could be corrected in
>> the future?
> 
> No, it's not intentional that pgpool-II 3.2.4 requires
> pgpool_regclass() (if it is so). Will look into this.

It turned out that it's a bug with 3.2.4. Attached patch should fix
the problem. Please try it.
--
Tatsuo Ishii
SRA OSS, Inc. Japan
English: http://www.sraoss.co.jp/index_en.php
Japanese: http://www.sraoss.co.jp
-------------- next part --------------
A non-text attachment was scrubbed...
Name: pool_select_walker.c.patch
Type: text/x-patch
Size: 716 bytes
Desc: not available
URL: <http://www.sraoss.jp/pipermail/pgpool-general/attachments/20130514/f63c0228/attachment.bin>


More information about the pgpool-general mailing list