[OpenAFS] Re: [OpenAFS] Re: [OpenAFS] Re: [OpenAFS] cache manager locked under heavy load?

Alena Manova nymano@seznam.cz
Mon, 08 Feb 2010 03:56:26 +0100 (CET)


> ------------ Původní zpráva ------------
> Od: Derrick Brashear <shadow@gmail.com>
> Předmět: Re: [OpenAFS] Re: [OpenAFS] Re: [OpenAFS] cache manager =
locked under
> heavy load?
> Datum: 08.2.2010 03:41:38
> ----------------------------------------
> 2010/2/7 Alena Manova <nymano@seznam.cz>:
> >> ------------ Původní zpráva ------------
> >> Od: Andrew Deason <adeason@sinenomine.net>
> >> Předmět: Re: [OpenAFS] Re: [OpenAFS] cache manager locked unde=
r heavy load?
> >> Datum: 08.2.2010 02:09:49
> >> ----------------------------------------
> >> On Mon, 08 Feb 2010 00:01:03 +0100 (CET)
> >> Alena Manova <nymano@seznam.cz> wrote:
> >>
> >> > I am running one fileserver process. should I run more? how to g=
et
> >> > the optimal number?
> >> [...]
> >> > Instance fs, (type is fs) currently running normally.
> >> >     Auxiliary status is: file server running.
> >> >     Process last started at Sun Feb  7 04:00:05 2010 (2 pr=
oc starts)
> >> >     Command 1 is '/usr/afs/bin/fileserver'
> >> >     Command 2 is '/usr/afs/bin/volserver'
> >> >     Command 3 is '/usr/afs/bin/salvager'
> >>
> >> The number of threads you're running with is usually given by the =
-p
> >> option to the fileserver. You are running with the default, which =
I
> >> think is around 9. Try adding the argument "-p 128" (no quotes) to=
 the
> >> parameters passed to the fileserver, and see if it affects your pr=
oblem
> >> at all.
> >>
> >
> > ok, thanks.
> >
> > I changed the fileserver to be started with the option -L which mea=
ns "large"
> and implies -p 128.
> >
> > however it didn't help solve my cache manager issue.
> 
> nor will it.
> >
> > I also followed the vCache performance analysis described by Jason =
but
> regardless the -stats parameter (even tried with value 1000000) I can=
't get
> better hits/misses ratio than about 50% (which is much worse than rec=
ommended
> 1%).
> 
> -stats is not -stat.
> 
> > how about the 'dynamic-vcaches' feature? it is enabled by default s=
o is the
> -stats parameter still relevant?
> 
> 
> > I tried to use the -fakestat-all parameter and it seems it can impr=
ove the
> ratio to 4%. but the issue persists (with same scale).
> >
> > currently I am running the cache manager with -dynroot -fakestat-al=
l -volumes
> 256 -daemons 6 -stats 300000 and the fileserver with 128 LWPs. but th=
e issue is
> still there:
> 
> as above. with with -stat rather than -stats
> 
> 

sorry, typo in my mail but not in my config - afsd would even not start with that misspelled parameter...

>