Bug #3226 | User variables in insert statements don't replicate properly | ||
---|---|---|---|
Submitted: | 18 Mar 2004 9:46 | Modified: | 25 Mar 2004 12:08 |
Reporter: | Gary Thornock | Email Updates: | |
Status: | Not a Bug | Impact on me: | |
Category: | MySQL Server: Replication | Severity: | S2 (Serious) |
Version: | 4.1 and previous | OS: | Linux (Linux) |
Assigned to: | CPU Architecture: | Any |
[18 Mar 2004 9:46]
Gary Thornock
[25 Mar 2004 10:37]
MySQL Verification Team
Worked for me with both 4.0.19 and 4.1.2: /*!40019 SET @@session.max_insert_delayed_threads=0*/; # at 4 #040325 20:39:02 server id 1 log_pos 4 Start: binlog v 3, server v 4.1.2-alpha-log created 040325 20:39:02 at startup # at 79 #040325 20:39:50 server id 1 log_pos 79 User_var SET @sinisa:=112; # at 126 #040325 20:39:50 server id 1 log_pos 126 Query thread_id=1 exec_time=0 error_code=0 use telcent; SET TIMESTAMP=1080239990; insert into nazivi values (@sinisa,"Yes");
[25 Mar 2004 12:08]
Guilhem Bichot
Correction to the previous answer. It does not work in 4.0, which is a documented bug. User variables replicate fine starting from 4.1.0 (so master and slave must be _both_ 4.1.0 or 4.1.1). So it's normal that it does not work when using MySQL "previous to" 4.1. If your master and slave are both 4.1.0 or 4.1.1, please re-open the bug report. Thank you.