| Bug #25705 | Federated: CREATE SERVER logging | ||
|---|---|---|---|
| Submitted: | 18 Jan 2007 20:08 | Modified: | 13 Oct 2009 0:57 |
| Reporter: | Peter Gulutzan | Email Updates: | |
| Status: | Won't fix | Impact on me: | |
| Category: | MySQL Server: Federated storage engine | Severity: | S4 (Feature request) |
| Version: | 5.1.15-beta-debug-log | OS: | Linux (SUSE 10.0 / 64-bit) |
| Assigned to: | Luis Soares | CPU Architecture: | Any |
| Tags: | disabled | ||
[18 Jan 2007 20:08]
Peter Gulutzan
[18 Jan 2007 22:55]
MySQL Verification Team
Thank you for the bug report.
[19 Jan 2009 15:24]
Luis Soares
Hi, while dealing with BUG#13684 I have found that the create server statement is actually binlogged when using *row* format, but not on *mixed* nor *statement*. You can verify this by doing the following on the latest 5.1 tree (I used revno: 2710,revision-id: joerg@mysql.com-20090115181125-29qdo615v9vkla0t): mysql> SET SESSION BINLOG_FORMAT='statement'; Query OK, 0 rows affected (0.00 sec) mysql> create server s_statement foreign data wrapper mysql options (user 'Remote', Host '192.168.1.103', database 'test'); Query OK, 1 row affected (0.00 sec) mysql> show binlog events; +------------------+-----+-------------+-----------+-------------+---------------------------------------------+ | Log_name | Pos | Event_type | Server_id | End_log_pos | Info | +------------------+-----+-------------+-----------+-------------+---------------------------------------------+ | mysql-bin.000001 | 4 | Format_desc | 1 | 106 | Server ver: 5.1.32-debug-log, Binlog ver: 4 | +------------------+-----+-------------+-----------+-------------+---------------------------------------------+ 1 row in set (0.00 sec) mysql> reset master; Query OK, 0 rows affected (0.01 sec) mysql> SET SESSION BINLOG_FORMAT='mixed'; Query OK, 0 rows affected (0.00 sec) mysql> create server s_mixed foreign data wrapper mysql options (user 'Remote', Host '192.168.1.103', database 'test'); Query OK, 1 row affected (0.00 sec) mysql> show binlog events; +------------------+-----+-------------+-----------+-------------+---------------------------------------------+ | Log_name | Pos | Event_type | Server_id | End_log_pos | Info | +------------------+-----+-------------+-----------+-------------+---------------------------------------------+ | mysql-bin.000001 | 4 | Format_desc | 1 | 106 | Server ver: 5.1.32-debug-log, Binlog ver: 4 | +------------------+-----+-------------+-----------+-------------+---------------------------------------------+ 1 row in set (0.00 sec) mysql> reset master; Query OK, 0 rows affected (0.00 sec) mysql> SET SESSION BINLOG_FORMAT='row'; Query OK, 0 rows affected (0.00 sec) mysql> create server s_row foreign data wrapper mysql options (user 'Remote', Host '192.168.1.103', database 'test'); Query OK, 1 row affected (0.00 sec) mysql> show binlog events; +------------------+-----+-------------+-----------+-------------+---------------------------------------------+ | Log_name | Pos | Event_type | Server_id | End_log_pos | Info | +------------------+-----+-------------+-----------+-------------+---------------------------------------------+ | mysql-bin.000001 | 4 | Format_desc | 1 | 106 | Server ver: 5.1.32-debug-log, Binlog ver: 4 | | mysql-bin.000001 | 106 | Query | 1 | 170 | BEGIN | | mysql-bin.000001 | 170 | Table_map | 1 | 242 | table_id: 12 (mysql.servers) | | mysql-bin.000001 | 242 | Write_rows | 1 | 319 | table_id: 12 flags: STMT_END_F | | mysql-bin.000001 | 319 | Query | 1 | 384 | COMMIT | +------------------+-----+-------------+-----------+-------------+---------------------------------------------+ 5 rows in set (0.00 sec) Using mysqlbinlog on the resulting binlog file I can check that the event is the actual insert in the mysql.servers table: ../bin/mysqlbinlog -vv mysql-bin.000001 (...) # at 242 #090119 16:11:23 server id 1 end_log_pos 242 Table_map: `mysql`.`servers` mapped to number 12 #090119 16:11:23 server id 1 end_log_pos 319 Write_rows: table id 12 flags: STMT_END_F BINLOG ' G5h0SRMBAAAASAAAAPIAAAAAAAwAAAAAAAAABW15c3FsAAdzZXJ2ZXJzAAn+/v7+/gP+/v4Q/sD+ wP7A/sD+wP7A/sD+wAAA G5h0SRcBAAAATQAAAD8BAAAQAAwAAAAAAAEACf//AP4Fc19yb3cNMTkyLjE2OC4xLjEwMwR0ZXN0 BlJlbW90ZQAAAAAAAAVteXNxbAA= '/*!*/; ### INSERT INTO mysql.servers ### SET ### @1='s_row' /* STRING(192) meta=65216 nullable=0 is_null=0 */ ### @2='192.168.1.103' /* STRING(192) meta=65216 nullable=0 is_null=0 */ ### @3='test' /* STRING(192) meta=65216 nullable=0 is_null=0 */ ### @4='Remote' /* STRING(192) meta=65216 nullable=0 is_null=0 */ ### @5='' /* STRING(192) meta=65216 nullable=0 is_null=0 */ ### @6=0 /* INT meta=0 nullable=0 is_null=0 */ ### @7='' /* STRING(192) meta=65216 nullable=0 is_null=0 */ ### @8='mysql' /* STRING(192) meta=65216 nullable=0 is_null=0 */ ### @9='' /* STRING(192) meta=65216 nullable=0 is_null=0 */ # at 319 #090119 16:11:23 server id 1 end_log_pos 384 Query thread_id=1 exec_time=0 error_code=0 SET TIMESTAMP=1232377883/*!*/; COMMIT /*!*/; DELIMITER ; # End of log file ROLLBACK /* added by mysqlbinlog */; /*!50003 SET COMPLETION_TYPE=@OLD_COMPLETION_TYPE*/; Regards, Luís
[9 Feb 2009 13:44]
Luis Soares
BUG#13684 is (somewhat) related to this one.
[9 Feb 2009 15:54]
Bugs System
A patch for this bug has been committed. After review, it may be pushed to the relevant source trees for release in the next version. You can access the patch from: http://lists.mysql.com/commits/65629 2768 Luis Soares 2009-02-09 BUG#25705: Federated: CREATE SERVER logging BUG#13684: SP: DROP PROCEDURE|FUNCTION IF EXISTS not binlogged if routine does not exist From BUG#25705, one finds that the drop/create server is not binlogged in mixed or statement mode replication. Instead it gets logged only in row based replication as rows inserted/deleted into/from mysql.servers. From BUG#13684, one finds that the drop server if exists is not being binlogged also. This patch addresses these issues by binlogging CREATE/DROP server even if the "IF EXISTS" clause is present. Furthermore, it will always log create/drop server as statements as these are referred in the manual to be "DDL": * http://dev.mysql.com/doc/refman/5.1/en/drop-server.html * http://dev.mysql.com/doc/refman/5.1/en/create-server.html
[5 Jun 2009 15:43]
Bugs System
A patch for this bug has been committed. After review, it may be pushed to the relevant source trees for release in the next version. You can access the patch from: http://lists.mysql.com/commits/75721 2917 Luis Soares 2009-06-05 BUG#25705: Federated: CREATE SERVER logging From BUG#25705, one finds that the CREATE/ALTER/DROP SERVER is not binlogged in mixed or statement mode replication. Instead it gets logged only in row based replication when rows are inserted/updated/deleted into/from mysql.servers. From BUG 13684, one finds that the drop server if exists is not being binlogged also. This patch addresses these issues by binlogging CREATE/ALTER/DROP SERVER even if the server does not exist but "IF EXISTS" clause was issued. Furthermore, it will always log CREATE/ALTER/DROP server as statements. @ mysql-test/suite/binlog/t/binlog_create_server.test Added test case that performs several checks involving CREATE/ALTER/DROP SERVER binlogging. @ sql/sql_parse.cc Changed error handling in sql_parse for drop server as now the drop_server call may return also errors related to binlogging. Also, extends error reporting to handle previous unhandled errors returned by drop server. @ sql/sql_servers.cc Deployed calls to THD::binlog_query in create_server, alter_server drop_server functions.
[8 Oct 2009 12:44]
Lars Thalmann
Push to Betony or Celosia unless Staale objects.
