In the last episode (Apr 17), Bernd Walter said: > On Fri, Apr 17, 2009 at 06:28:29PM +0200, Marius NĂ¼nnerich wrote: > > On Fri, Apr 17, 2009 at 16:18, Bernd Walter <ticso_at_cicely7.cicely.de> wrote: > > > On Fri, Apr 17, 2009 at 02:50:24PM +0200, Alexander Leidinger wrote: > > >> I monitored (by hand) a while the sysctls > > >> kstat.zfs.misc.arcstats.size and kstat.zfs.misc.arcstats.hdr_size. > > >> Both grow way higher (at some point I've seen more than 500M) than > > >> what I have configured in vfs.zfs.arc_max (40M). > > > > > > My understanding about this is the following: vfs.zfs.arc_min/max are > > > not used as min max values. They are used as high/low watermarks. If > > > arc is more than max the arc a thread is triggered to reduce the arc > > > cache until min, but in the meantime other threads can still grow arc > > > so there is a race between them. > > > > Hmm, if this is true the ARC size should go down to arc_min once it did > > grow past arc_max and no new data is coming along but I do not observe > > such a thing here. It simply stays near but below arc_max here all the > > time. I have only /home on ZFS with moderate load. > > I had a few ideas why this could be, but scanning complete sys showed no > point at all where arc_min is used. There are formular to set this value, > but that's all I find. zfs_arc_{min,max} are just tunables. The real variables arc_c_{min,max} get autosized and then capped to {min,max} in uts/common/fs/zfs/arc.c:arc_init() . -- Dan Nelson dnelson_at_allantgroup.comReceived on Fri Apr 17 2009 - 19:13:58 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:39:46 UTC