[pgpool-hackers: 21] Re: Road map to next minor version up

Bryan Varner bvarner at polarislabs.com
Mon Jan 9 23:43:23 JST 2012


Sure this.

This was created against V3_1_STABLE. I'm including it in the body of 
the email since attaching patches seems to be what keeps my messages 
from posting to this list.

Regards,
-Bryan



diff --git a/pool_proto_modules.c b/pool_proto_modules.c
index 586ab17..ae8065e 100644
--- a/pool_proto_modules.c
+++ b/pool_proto_modules.c
@@ -546,7 +546,6 @@ POOL_STATUS Execute(POOL_CONNECTION *frontend, 
POOL_CONNECTION_POOL *backend,
  	 * Decide where to send query
  	 */
  	session_context->query_context = query_context;
-	pool_where_to_send(query_context, query, node);

  	/* check if query is "COMMIT" or "ROLLBACK" */
  	commit = is_commit_or_rollback_query(node);
@@ -969,8 +968,6 @@ POOL_STATUS Bind(POOL_CONNECTION *frontend, 
POOL_CONNECTION_POOL *backend,
  	}

  	session_context->query_context = query_context;
-	pool_where_to_send(query_context, query_context->original_query,
-					   query_context->parse_tree);

  	if (pool_config->load_balance_mode && pool_is_writing_transaction())
  	{
@@ -1048,8 +1045,6 @@ POOL_STATUS Describe(POOL_CONNECTION *frontend, 
POOL_CONNECTION_POOL *backend,
  	}

  	session_context->query_context = query_context;
-	pool_where_to_send(query_context, query_context->original_query,
-					   query_context->parse_tree);

  	pool_debug("Describe: waiting for master completing the query");
  	if (pool_extended_send_and_wait(query_context, "D", len, contents, 1, 
MASTER_NODE_ID)
@@ -1117,7 +1112,6 @@ POOL_STATUS Close(POOL_CONNECTION *frontend, 
POOL_CONNECTION_POOL *backend,
  	}

  	session_context->query_context = query_context;
-	/* pool_where_to_send(query_context, query_context->original_query, 
query_context->parse_tree); */

  	pool_debug("Close: waiting for master completing the query");
  	if (pool_extended_send_and_wait(query_context, "C", len, contents, 1, 
MASTER_NODE_ID)



On 01/06/2012 09:20 PM, Tatsuo Ishii wrote:
> Can you please repost patches against git V3_1_STABLE or V3_0_STABLE
> in unified diff context diff format?
> --
> Tatsuo Ishii
> SRA OSS, Inc. Japan
> English: http://www.sraoss.co.jp/index_en.php
> Japanese: http://www.sraoss.co.jp
>
>>> Now that two major problems have been identified[1][2], I think we
>>> need to release minor version up of pgpool-II 3.1 and 3.0 stable
>>> trees.
>>>
>>> [1]
>>> http://www.pgpool.net/pipermail/pgpool-general/2012-January/000126.html
>>> [2]
>>> http://www.pgpool.net/pipermail/pgpool-general/2011-December/000121.html
>>>
>>> #1 seems to be already fixed according to users's reports.
>>> For #2, I believe Kitagawa is working on it.
>>>
>>> I think we should release 3.1.2 and 3.0.6 soon, hopefully by the end
>>> of January. Comments?
>>
>> Tatsuo, based on a conversation I had with you back in October on the
>> old pgpool-general, I've put together a patch for the issue of
>> pool_where_to_send() being invoked too often, causing application
>> failures when running in master_slave&  load_balance.
>>
>> I've tried to send the patch to this list three times in the last
>> week, and it's not going through.
>>
>> I'd really like to have this included in the next release if
>> possible. It's a show-stopper for our intended use of pgpool.
>>
>> Regards,
>> -Bryan Varner
>> _______________________________________________
>> pgpool-hackers mailing list
>> pgpool-hackers at pgpool.net
>> http://www.pgpool.net/mailman/listinfo/pgpool-hackers



More information about the pgpool-hackers mailing list