Joseph Qi
2014-Jan-28 09:53 UTC
[Ocfs2-devel] [RFD] ocfs2: poor performance on append write/punch hole
In ocfs2_file_aio_write, it will fall back to buffered IO when cannot complete direct IO request, e.g. append write and punch hole. And mostly it should firstly do space allocation and this will make much more journal to flush.>From test, we found the above two cases performs almost 70% lower thannormal case(allocate space first and overwrite). Furthermore, it spends more than 90% time on jbd2_journal_force_commit during aio write. So if anyone have an idea on how to tune the performance on these two cases?