[Bug 253968] 13.0-BETA4: UFS with journaled soft-updates is three times SLOWER than before

[bugzilla-noreply at freebsd.org]

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=253968

--- Comment #13 from Mark Millard <marklmi26-fbsd at yahoo.com> ---
Being unable to replicate anything similar to comment #4 and
comment #11 being a large increase in Inter-Process-Communication
and the rest of the rather-modern-FreeBSD examples being UFS SU
(no J) that do not show significant problems, then it would
appear that the UFS+SU+J (so with journaling) status of the
original report may be an essentially separate issue from other
fairly recent UFS I/O performance problems that seem to have
been addressed.

I do not plan to report other examples that try to isolate the
context vs. recent history. It seems that only journaled
experiments would likely be relevant to the epecific issue
being reported.

Sorry it took me so long to reach that conclusion.

The comment #11's IPC operation counts in doing a portsnap
extract should probably be a separate submittal.

I'm less clear on comment #4 vs. some context local issue
(say, file system fragmentation just to illustrate). But
if it can be replicated by others, it too should probably
have a separate submittal.

-- 
You are receiving this mail because:
You are the assignee for the bug.