Code review comment for lp:~gl-az/percona-server/ST-41544-5.5

On 07/05/2014 05:45 AM, Laurynas Biveinis wrote:
> - Is this bit really correct? (and the rest of binlog_stm_binlog)
> We shouldn't be affecting SBR replication of temp tables, should
> we?
>
> --- mysql-test/suite/binlog/r/binlog_stm_binlog.result 2011-02-23 11:54:58 +0000
> +++ mysql-test/suite/binlog/r/binlog_stm_binlog.result 2014-06-19 20:16:48 +0000
> @@ -635,9 +635,7 @@
> COERCIBILITY(s1) d3;
> DROP TEMPORARY TABLE tmp1;
> END
> -master-bin.000001 # Query # # use `bug39182`; CREATE TEMPORARY TABLE tmp1
> -SELECT * FROM t1 WHERE a LIKE CONCAT("%", NAME_CONST('s1',_utf8'test' COLLATE 'utf8_unicode_ci'), "%")
> -master-bin.000001 # Query # # use `bug39182`; DROP TEMPORARY TABLE `tmp1` /* generated by server */
> +master-bin.000001 # Query # # DROP TEMPORARY TABLE IF EXISTS `bug39182`.`tmp1` /* generated by server */
> DROP PROCEDURE p1;
> DROP TABLE t1;
> DROP DATABASE bug39182;
>
OK, yes, this is correct. I didn't catch it before but the issue is that
the test case name implies SBR, but in fact it is specifically testing
MBR instead, thus these result changes are correct. I suppose that could
be a bug in itself that the test is binlog_stm_binlog.test. I think it
would probably be a good idea to add/record a binlog_mix_binlog.test and
fix the _stm_ to actualy test SBR.

--
George O. Lorch III
Software Engineer, Percona
+1-888-401-3401 x542 US/Arizona (GMT -7)
skype: george.ormond.lorch.iii