system slowdown - vnode related

[Masachika ISHIZUKA]

>>>   I have still vnodes problem in 4.8-stable with /sys/kern/vfs_subr.c
>>> 1.249.2.30.
> 
> Ishizuka-san, could you possibly try the following command line
> repeatedly while slowdown is being observed?
> 
> 	% vmstat -m | grep '^ *vfscache'
> 
> If the third number of its output is approaching or hitting the fourth,
> the chances are your kernel is running out of memory for namecache,
> which was actually the case on my machines.

  Hi, nagao-san.

  I stopped 310.locate of weekly cron and the slow down is not occurred so
often. The slow down was just occurred as follows with a dual Xeon machine
(Xeon 2.4GHz x 2, 2 giga byte rams, 4.8-Stable with SMP and HTT option).

% sysctl -a|grep vnodes
kern.maxvnodes: 140000
kern.minvnodes: 33722
debug.numvnodes: 140025
debug.wantfreevnodes: 25
debug.freevnodes: 76
% vmstat -m | grep '^ *vfscache'
     vfscache818445 52184K  72819K102400K 19758622    0     0  64,128,256,512K

  It seems that the third number is smaller enough than fourth.
  I typed 'sysctl kern.maxvnodes=150000' and the machine is recovered.

-- 
ishizuka at ish.org