Re: Reducing downtime on system critical services
On Tue, Apr 04, 2000 at 02:17:35AM -0700, Kevin Dalley wrote:
> Ben Collins <bcollins@debian.org> writes:
>
> > On Mon, Apr 03, 2000 at 11:38:07AM -0700, Brent Fulgham wrote:
> > > > now that you mention it, that would be the best solution. (and is)
> > > > i just never thought of that. but this proves the point of community
> > > > developing something, it brings out choices that usually are
> > > > overlooked.
> > >
> > > The only problem with this approach is in cases where the running
> > > program makes use of the file you want to change. In that case,
> > > you may screw things up by swapping a configuration file out from
> > > underneath the program if it's still running.
> >
> > Yeah, and those programs should special case themselves (and policy should
> > allow it).
> >
> > I think those types are rare and few between though (usually if the
> > program monitors the config file state actively, the changes should not be
> > so that it would cause problems). Arguably, such programs are broken to
> > start with. I don't think I would want anything that immeiately recognized
> > my config file changes without telling it to (with a simple SIGHUP, like
> > inetd and such do).
>
> Is cron broken? Each minute, it checks its spool directory an
> /etc/cron.d.
Those aren't config files, they are scripts to be run. The files in the
spool directory are also checked (atleast normally) by crontab before
making them active. So there is some measure of safety in this setup.
On the other hand, if slapd actively checked it's config file, and there
was a problem in it, then slapd would quite (cron doesn't die because of a
bad crontab or script in the /etc/cron.* directories).
I never said there weren't special cases :)
--
-----------=======-=-======-=========-----------=====------------=-=------
/ Ben Collins -- ...on that fantastic voyage... -- Debian GNU/Linux \
` bcollins@debian.org -- bcollins@openldap.org -- bmc@visi.net '
`---=========------=======-------------=-=-----=-===-======-------=--=---'
Reply to: