https://bugs.kde.org/show_bug.cgi?id=362055
Harald Sitter <sit...@kde.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |sit...@kde.org
Status|UNCONFIRMED |NEEDSINFO
Resolution|--- |WAITINGFORINFO
--- Comment #1 from Harald Sitter <sit...@kde.org> ---
I am not sure what that has to do with the default configuration I am afraid. I
mean, you can certainly fine tune that for your system, but for general purpose
I'd trust the defaults are useful.
If your RAM and your swap is full you are out of memory at which point you have
reached maximum spawness regardless of your swapiness and under continued load
the kernel will either be 100% IO waiting to swap/unswap events or attempt to
shoot randomly in the crowd to OOM kill a process that matters (which, unless
it is in make's parent group, won't happen).
What you describe sounds more like RAM exhaustion leading to swapping leading
to disk IO exhaustion leading to the system being sluggish to the point of
perceived freeze. Since you also see it during baloo startup (at which point
neither min_free_kbytes nor swapiness ought to matter) it definitely looks more
like general IO exhaustion. Filing a bug against baloo would probably be a good
first step of locking down the issue.
First please post the output of
`cat /sys/block/sda/queue/scheduler`
--
You are receiving this mail because:
You are watching all bug changes.
