Am Donnerstag, 17. März 2005 02:24 schrieb Jeff Roberson: > On Thu, 17 Mar 2005, Emanuel Strobl wrote: > > Am Mittwoch, 16. März 2005 07:16 schrieb Sergey Matveychuk: > > > Jeff Roberson wrote: > > > >>just ffs: 1:47 min > > > >>nullfs: 1:43 min (oops!:) > > > >>nullfs in jail: 12:12 min > > > >> > > > >>Almost ten times degradation. > > > > > > > > Can you try this again on current? I just committed some changes to > > > > nullfs which vastly simplify the locking. I don't see anything > > > > inherent in the code that should cause such a slowdown. > > > > > > The problem has gone. Thanks Jeff! > > > Don't forget MFC it some time :) > > > > Highly appreciated for 5.4!!! > > I can't MFC this for 5.4, it requires too much of the new vfs architecture > to properly function. I'll mail re_at_ to see what their opinion is for 5.5. Oic. I haven't tracked -current since christmas so I'd like to ask if you know anything new about file backed md performance. I wanted to use that for jail systems but performance was really bad. Does your great work also improve md behaviour? Thank you very much, -Harry
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:30 UTC