Re: NFS is going down, et al [was: pidfiles aka. Re: [CVS] unix socket support added]

[Christian Hopp]

On 6 Aug 2002, Jan-Henrik Haukeland wrote:

> Christian Hopp <address@hidden> writes:
>
> > The alarm signal is still evaluated... but later!  So as I told you
> > a NFS halted processes can't be woken up by anything.
>
> Ouch! I would love to know the inner technical reason for this. I
> guess I will have to start reading up on RPC and NFS.

I think that's gonna take a life time. (-:

> > > You're in an excellent position to test this then :) I'm beting you a
> > > bottle of beer that alarm will work.
> >
> > Our address is on our web page. (-: But for me something
> > non-alcoholic, please!
>
> Beer or nothing I'm afraid :) So we can't expect to see you in the
> crowd, the next Beer festival in October then?

I don't think so... anyways I am quite busy in October.  But maybe you
can donate the beer to some kind of aid organization? (-:


(...)

> > I think I can do it.  I already took a look into the /proc access in
> > Linux and Solaris.  It's gonna be quite OS dependent.  I will give it
> > a generic frame work to access it OS independent, or can anyone
> > recommend me a good lib for it, but the backend is different on
> > esp. Linux and Solaris.
>
> I was planing to look into libgtop which seems to be a usefull library
> for this.
>
>   http://www.redhat.com/swr/i386/libgtop-devel-1.0.12-8.i386.html
>
> Maybe this library will give you a good start? It's GPL and I was
> further planing to do some creative cut-and-past to create a mini
> package for usage in monit. The reason is that I don't want monit to
> depend on axillary libraries and users should _not_ have to download
> such libraries separately. One of the best things with monit (IMO) is
> that it's a compact little program without a bunch of pluggins, extra
> libraries and stuff - it's just one program with a good comprehensive
> configuration file.

Libgtop is no help to me.  The source of the procps tools is more
interesting.  And for Solaris there is nothing better then proc's man
page. By the way, Solaris has a really good and developer friendly
procfs.

I have already solaris code working to find out its status
(cpu+mem+zombie?).  For Linux I just got "zombie?"  working.  Linux's
proc is a real pain in the ****.  Maybe it's user friendly.  But if
you wanna read it reliable with C that its awful.  Linux also does not
provide any possibility to measure the "cpu percentage" of a task
directly.  So I gonna measure each cycle and calculate the difference
between the cylces.  Linux uses it's obscure "jiffies".  As far as I
know right now a jiffie is something between nothing and forever
(precise "usually 1/100 s except for alpha systems there it is 1/1024
seconds).  But it's going on.  I gonna try to keep the information to
a little key values (mem, cpu incl. children, status ->[ zombie,
stopped, halted, running]).

Bye,

Christian


-- 
Christian Hopp                                email: address@hidden
Institut für Elektrische Informationstechnik             fon: +49-5323-72-2113
Technische Universität Clausthal                         fax: +49-5323-72-3197
  pgpkey: https://www.iei.tu-clausthal.de/pgp-keys/chopp.key.asc  (2001-11-22)