| Bug #27044 | replicated with unique field ndb table allows duplicate key inserts | ||
|---|---|---|---|
| Submitted: | 12 Mar 2007 12:47 | Modified: | 2 Jun 2007 16:35 |
| Reporter: | Andrei Elkin | ||
| Status: | Closed | ||
| Category: | Server: ClusterRep | Severity: | S3 (Non-critical) |
| Version: | 5.1.17 | OS: | Any |
| Assigned to: | Andrei Elkin | Target Version: | |
[12 Mar 2007 20:11]
Andrei Elkin
Checked other engine to conclude that ndb only is vulnerable. The patch is coming soon.
[13 Mar 2007 11:20]
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/21790 ChangeSet@1.2472, 2007-03-13 12:18:57+02:00, aelkin@dsl-kpogw6-fe8ddc00-181.dhcp.inet.fi +3 -0 Bug#27044 replicated with unique field ndb table allows duplkey inserts The bug in that slave version of a table with unique field still was able to execute INSERT query as replace whereas it's impossible on master. The reason of this artifact is lack of resetting of extra() [see files' comments]
[13 Mar 2007 11:20]
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/21791 ChangeSet@1.2472, 2007-03-13 11:38:06+02:00, aelkin@dsl-kpogw6-fe8ddc00-181.dhcp.inet.fi +3 -0 Bug#27044 replicated with unique field ndb table allows duplkey inserts The bug in that slave version of a table with unique field still was able to execute INSERT query as replace whereas it's impossible on master. The reason of this artifact is lack of resetting of extra() [see files' comments]
[15 Mar 2007 19:36]
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/22043 ChangeSet@1.2472, 2007-03-15 20:35:36+02:00, aelkin@dsl-hkibrasgw1-ff1ec100-106.dhcp.inet.fi +3 -0 Bug#27044 replicated with unique field ndb table allows duplkey inserts The bug in that slave version of a table with unique field still was able to execute INSERT query as replace whereas it's impossible on master. The reason of this artifact is wrong usage of ndb->extra:s. Fixed with leaving in do_before only table->file->extra(HA_EXTRA_WRITE_CAN_REPLACE); and reseting in do_after.
[19 Mar 2007 9:36]
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/22233 ChangeSet@1.2472, 2007-03-19 10:35:20+02:00, aelkin@dsl-hkibrasgw1-ff1ec100-106.dhcp.inet.fi +3 -0 Bug#27044 replicated with unique field ndb table allows duplkey inserts The bug in that slave version of a table with unique field still was able to execute INSERT query as replace whereas it's impossible on master. The reason of this artifact is wrong usage of ndb->extra:s. Fixed with resetting flags in do_before. There is open issue with symmetrical resetting table->file->extra(HA_EXTRA_NO_IGNORE_NO_KEY) which i had to omit due to the issue reported as bug#27077
[22 Mar 2007 13:04]
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/22625 ChangeSet@1.2447, 2007-03-22 13:17:08+01:00, tomas@whalegate.ndb.mysql.com +2 -0 BUG#27044 bug fix of 27320 fixes this, added test case
[2 May 2007 11:44]
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/25891 ChangeSet@1.2472, 2007-05-02 12:44:26+03:00, aelkin@dsl-hkibras1-ff18c300-219.dhcp.inet.fi +3 -0 Bug#27044 replicated with unique field ndb table allows duplkey inserts The bug in that slave version of a table with unique field still was able to execute INSERT query as replace whereas it's impossible on master. The reason of this artifact is wrong usage of ndb->extra:s. Fixed with resetting flags at do_after. There is open issue with symmetrical resetting table->file->extra(HA_EXTRA_NO_IGNORE_NO_KEY) which i had to hand to bug#27077.
[2 May 2007 16:27]
Andrei Elkin
The last committed patch is for review.
[27 May 2007 14:20]
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/27406 ChangeSet@1.2472, 2007-05-27 15:20:20+03:00, aelkin@dsl-hkibras1-ff5dc300-70.dhcp.inet.fi +1 -0 Bug#27044 replicated with unique field ndb table allows duplkey inserts The bug in that slave version of a table with unique field still was able to execute INSERT query as replace whereas it's impossible on master. The reason of this artifact is wrong usage of ndb->extra:s. Fixed with resetting flags at do_after. There is open issue with symmetrical resetting table->file->extra(HA_EXTRA_NO_IGNORE_NO_KEY) which i had to hand to bug#27077. The test for the current bug was committed in a cset for bug#27320.
[29 May 2007 15:36]
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/27573 ChangeSet@1.2472, 2007-05-29 16:36:05+03:00, aelkin@dsl-hkibras1-ff5dc300-70.dhcp.inet.fi +1 -0 Bug#27044 replicated with unique field ndb table allows duplkey inserts The bug in that slave version of a table with unique field still was able to execute INSERT query as replace whereas it's impossible on master. The reason of this artifact is wrong usage of ndb->extra:s. Fixed with resetting flags at do_after. There is open issue with symmetrical resetting table->file->extra(HA_EXTRA_NO_IGNORE_NO_KEY) which i had to hand to bug#27077. The test for the current bug was committed in a cset for bug#27320.
[29 May 2007 16:08]
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/27582 ChangeSet@1.2472, 2007-05-29 15:53:03+03:00, aelkin@dsl-hkibras1-ff5dc300-70.dhcp.inet.fi +1 -0 Bug#27044 replicated with unique field ndb table allows duplkey inserts The bug in that slave version of a table with unique field still was able to execute INSERT query as replace whereas it's impossible on master. The reason of this artifact is wrong usage of ndb->extra:s. Fixed with resetting flags at do_after. There is open issue with symmetrical resetting table->file->extra(HA_EXTRA_NO_IGNORE_NO_KEY) which i had to hand to bug#27077. The test for the current bug was committed in a cset for bug#27320.
[1 Jun 2007 21:21]
Bugs System
Pushed into 5.1.20-beta
[2 Jun 2007 16:35]
Jon Stephens
Thank you for your bug report. This issue has been committed to our source repository of
that product and will be incorporated into the next release.
If necessary, you can access the source repository and build the latest available
version, including the bug fix. More information about accessing the source trees is
available at
http://dev.mysql.com/doc/en/installing-source.html
Documented bugfix in 5.1.20 changelog.
Description: starting cluster for replication with mysql-test-run.pl --start-and-exit --debug --mysqld=--binlog-format=row t/rpl_ndb_basic.test 1. create table with primary key 2. populate 3. trying to insert on master -> errror with dup key However, on slave: mysql> show create table t3; +-------+-----------------------------------------------------------------+ | Table | Create Table | +-------+-----------------------------------------------------------------+ | t3 | CREATE TABLE `t3` ( `a` int(11) NOT NULL, `b` int(11) DEFAULT NULL, PRIMARY KEY (`a`) ) ENGINE=ndbcluster mysql> select * from t3; +---+------+ | a | b | +---+------+ | 3 | 3 | | 1 | 1 | | 2 | 2 | +---+------+ 3 rows in set (0.01 sec) the same mysql> insert into t1 values (1,0),(2,0),(3,0); completes (from the result file of How to repeat test) select * from t1; a b 3 0 1 0 2 0 I noticed that repeating attempts to insert into unique the table starts rejecting. How to repeat: mysql-test-run.pl --record --mysqld=--binlog-format=row t/rpl_row_ndb_bug.test cat <<. >t/rpl_row_ndb_bug.test -- source include/have_ndb.inc let $type= 'NDB' ; -- source include/have_binlog_format_row.inc -- source include/master-slave.inc connection master; --disable_warnings drop table if exists t1; --enable_warnings eval CREATE TABLE t1 (a int primary key, b int) ENGINE= $type; insert into t1 values (1,1),(2,2),(3,3); --error ER_DUP_KEY insert into t1 values (1,0),(2,0),(3,0); select * from t1; sync_slave_with_master; select * from t1; insert into t1 values (1,0),(2,0),(3,0); select * from t1; . Suggested fix: To start from looking inside log_event.cc's Write_rows_log_event::do_before_row_operations, do_after make manipulations with table handle.