Comment 3 for bug 1667307

I think that this bug is actually not a duplicate of lp:1594964, although both have the same referenced upstream report.

The mem leak in this report happens only on a multi-source slave, and leak happens in memory/sql/Log_event area. While the bug in lp:1594964 is reproducible in simple master-master scenario, and leak happens in memory/sql/thd::main_mem_root area.
Although both issues seem similar and maybe even related in the code, as some conditions are the same, though the additional conditions that trigger the leaks are different in both cases.