Bug #44346 Changed configuation of "cluster" is not reflected in the tests
Submitted: 17 Apr 2009 18:05
Reporter: Joerg Bruehe Email Updates:
Status: Verified Impact on me:
None 
Category:Tests: Server Severity:S3 (Non-critical)
Version:Cluster 6.2 and up OS:Any
Assigned to: CPU Architecture:Any

[17 Apr 2009 18:05] Joerg Bruehe
Description:
Test "ndb_storedproc_08" fails in all runs of cluster-7.0.5.
The same failure is in the logs of cluster-6.2.17, but was not reported back then.

=====
funcs_1.ndb_storedproc_08                [ fail ]
        Test ended at YYYY-MM-DD HH:MM:SS

CURRENT_TEST: funcs_1.ndb_storedproc_08
--- /PATH/mysql-test/suite/funcs_1/r/ndb_storedproc_08.result
+++ /PATH/mysql-test/suite/funcs_1/r/ndb_storedproc_08.reject
@@ -228,8 +228,8 @@
 SQL_MODE
 ROUTINE_COMMENT
 DEFINER        root@localhost
-CHARACTER_SET_CLIENT   latin1
-COLLATION_CONNECTION   latin1_swedish_ci
+CHARACTER_SET_CLIENT   utf8
+COLLATION_CONNECTION   utf8_general_ci
 DATABASE_COLLATION     latin1_swedish_ci
 SPECIFIC_NAME  check_testcase
 ROUTINE_CATALOG        NULL
@@ -251,8 +251,8 @@
 SQL_MODE
 ROUTINE_COMMENT
 DEFINER        root@localhost
-CHARACTER_SET_CLIENT   latin1
-COLLATION_CONNECTION   latin1_swedish_ci
+CHARACTER_SET_CLIENT   utf8
+COLLATION_CONNECTION   utf8_general_ci
 DATABASE_COLLATION     latin1_swedish_ci
 SPECIFIC_NAME  check_warnings
 ROUTINE_CATALOG        NULL
@@ -274,8 +274,8 @@
... (same difference repeated several times)
=====

How to repeat:
Run the tests.

Suggested fix:
Re-record the expected test result.
[17 Apr 2009 19:05] Joerg Bruehe
Same test failure for "memory_storedproc_08" and "myisam_storedproc_08".
[17 Apr 2009 19:13] Joerg Bruehe
... not to forget "innodb_storedproc_08".
[17 Apr 2009 19:21] Joerg Bruehe
And this test failure has the same root cause:

=====
funcs_1.storedproc                       [ fail ]
        Test ended at YYYY-MM-DD HH:MM:SS

CURRENT_TEST: funcs_1.storedproc
--- /PATH/mysql-test/suite/funcs_1/r/storedproc.result
+++ /PATH/mysql-test/suite/funcs_1/r/storedproc.reject
@@ -94,10 +94,10 @@
 ERROR 42000: Identifier name 'sp1_thisisaveryverylongname234872934_thisisaveryverylongnameabcde' is too long
 SHOW PROCEDURE STATUS;
 Db     Name    Type    Definer Modified        Created Security_type   Comment character_set_client    collat
ion_connection  Database Collation
-mtr    add_suppression PROCEDURE       root@localhost  <modified>      <created>       DEFINER         latin1
        latin1_swedish_ci       latin1_swedish_ci
-mtr    check_testcase  PROCEDURE       root@localhost  <modified>      <created>       DEFINER         latin1
        latin1_swedish_ci       latin1_swedish_ci
-mtr    check_warnings  PROCEDURE       root@localhost  <modified>      <created>       DEFINER         latin1
        latin1_swedish_ci       latin1_swedish_ci
-mtr    force_restart   PROCEDURE       root@localhost  <modified>      <created>       DEFINER         latin1
        latin1_swedish_ci       latin1_swedish_ci
+mtr    add_suppression PROCEDURE       root@localhost  <modified>      <created>       DEFINER         utf8    utf8_general_ci latin1_swedish_ci
+mtr    check_testcase  PROCEDURE       root@localhost  <modified>      <created>       DEFINER         utf8    utf8_general_ci latin1_swedish_ci
+mtr    check_warnings  PROCEDURE       root@localhost  <modified>      <created>       DEFINER         utf8    utf8_general_ci latin1_swedish_ci
+mtr    force_restart   PROCEDURE       root@localhost  <modified>      <created>       DEFINER         utf8    utf8_general_ci latin1_swedish_ci
 DROP PROCEDURE IF EXISTS sp1;
 CREATE PROCEDURE sp1( f1 BINARY )
 LANGUAGE SQL DETERMINISTIC SQL SECURITY INVOKER COMMENT 'this is simple'
@@ -113,10 +113,10 @@
... (more similar differences)
=====
[4 May 2009 9:56] Joerg Bruehe
For the record, and nor surprise: Same issue in the build of 6.3.24