| Bug #934 | LOAD DATA FROM MASTER causes slave segfault when master runs without --log-bin | ||
|---|---|---|---|
| Submitted: | 27 Jul 2003 17:06 | Modified: | 28 Jul 2003 8:36 |
| Reporter: | Guilhem Bichot | ||
| Status: | Closed | ||
| Category: | Server: Replication | Severity: | S3 (Non-critical) |
| Version: | 4.0 | OS: | Any (all) |
| Assigned to: | Guilhem Bichot | Target Version: | |
[28 Jul 2003 8:36]
Guilhem Bichot
Thank you for your bug report. This issue has been fixed in the latest
development tree for that product. You can find more information about
accessing our development trees at
http://www.mysql.com/doc/en/Installing_source_tree.html
ChangeSet@1.1549, 2003-07-28 15:33:46+02:00, guilhem@mysql.com
Description: See how-to-repeat. How to repeat: Run the master without --log-bin, and on the slave (with the slave threads started or not) do: LOAD DATA FROM MASTER; It segfaults. Suggested fix: Here's a patch against sql/repl_failsafe.cc which solves it (I will push this patch to our development tree soon). *** 860,866 **** of LOAD DATA FROM MASTER - no reason to forbid it, really, although it does not make much sense for the user to do it */ ! if (row[0] && row[1]) { strmake(active_mi->master_log_name, row[0], sizeof(active_mi->master_log_name)); --- 860,866 ---- of LOAD DATA FROM MASTER - no reason to forbid it, really, although it does not make much sense for the user to do it */ ! if (row && row[0] && row[1]) { strmake(active_mi->master_log_name, row[0], sizeof(active_mi->master_log_name));