Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 15 Sep 2010 01:04:12 -0700
From:      Jeremy Chadwick <freebsd@jdc.parodius.com>
To:        perryh@pluto.rain.com
Cc:        ports@freebsd.org, swhetzel@gmail.com, itetcu@freebsd.org, dougb@freebsd.org, nork@freebsd.org
Subject:   Re: [ports/net/isc-dhcp*] Don't stop DHCP related daemons
Message-ID:  <20100915080412.GA54898@icarus.home.lan>
In-Reply-To: <4c90738f.W4uzxD1k7guRTp%2Bn%perryh@pluto.rain.com>
References:  <20100911222902.bb57444a.nork@FreeBSD.org> <20100911173359.68d71af6@it.buh.tecnik93.com> <AANLkTin0D13%2B9JQJ6wVD7_48Es6rvUJtu=rAx6M1xdjx@mail.gmail.com> <4C901FD0.5000304@FreeBSD.org> <4c90738f.W4uzxD1k7guRTp%2Bn%perryh@pluto.rain.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Sep 15, 2010 at 12:19:43AM -0700, perryh@pluto.rain.com wrote:
> Doug Barton <dougb@freebsd.org> wrote:
> 
> > >>>        ... Is it really absolutely necessary
> > >>>        to stop a service before it's files go away?
> >
> > IMO the only time the ports infrastructure itself should do this
> > is if it isn't possible to pkg_delete the port cleanly if it's
> > running. For example, if there is a file being held open that
> > cannot be deleted unless the service is stopped ...
> 
> Which should be an exceedingly rare circumstance, since the fact
> of some process(es) having a file open will not ordinarily prevent
> the removal of any or all directory entries pointing to it.  The
> inode and disk space won't actually be released until the last
> such process closes the file, but another file could be created
> having the same pathname as the one whose prior directory entry
> was removed.

...which also makes the assumption the daemon doesn't do stat(2) or
similar to verify a file exists, or does a multitude of other things
that might act on a filesystem but not an open descriptor.

We can sit here discussing what a daemon might do or not do until we're
blue in the face, which is why I tend to side with Doug's argument that
these sorts of tasks (stopping daemons, starting daemons, etc.) should
be left to the administrator.  I sympathise with the OP, but as I stated
previously: what kind of administrator upgrades software, especially a
daemon, then doesn't test/check to make sure everything's running
correctly after the upgrade?  I would rather we not try to solve a
borderline social problem with software.

But, also like Doug, I see the validity in the need for an automated
upgrade infrastructure/framework that can provide daemon auto-stop and
auto-start (I strongly oppose the latter) if desired.  I just don't know
how feasible that is, or if it's worth the time.

-- 
| Jeremy Chadwick                                   jdc@parodius.com |
| Parodius Networking                       http://www.parodius.com/ |
| UNIX Systems Administrator                  Mountain View, CA, USA |
| Making life hard for others since 1977.              PGP: 4BD6C0CB |




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20100915080412.GA54898>