Re: RFC: Alternate patch to have true new-style rc.d scripts in ports(without touching localpkg)

From: Jamie Bowden <ragnar_at_sysabend.org>
Date: Tue, 17 Aug 2004 10:13:53 -0700 (PDT)
On Tue, 17 Aug 2004, Andre Oppermann wrote:

> Brad Knowles wrote:
> > At 3:11 PM +0200 2004-08-17, sthaug_at_nethelp.no wrote:
> >
> >>  - Why cannot /usr/local/etc/rc.d be used with rcorder if /etc/rc.d/local
> >>  is okay?
> >
> >     You can't guarantee that /usr/local is on the same filesystem as
> > /etc (and available early in the boot process), and while you can't
> > guarantee that /etc/rc.d/local is also on the same filesytem, it's a lot
> > easier for some people to guarantee.
> >
> >     Moreover, many sites may mount /usr read-only, or /usr/local non-suid.
>
> So?  How does this prevent rc scripts being used from /usr or /usr/local?

Imagine /usr/local is NFS mounted from one machine across multiple in your
network.  Now imagine each machine needs a local set of individual startup
scripts for certain bits of software specific to it.  I have seen this
setup more than once.  On Irix and Solaris (and other SysV derived
systems), this isn't a problem, you just put the startup scripts for the
software you want to run in /etc/rcN.d/SNNscriptname, and you control the
order of startup and which particiular pieces of software startup.  Od you
also have the option of KNNscriptname for processes that must shut down
cleanly when changing run levels.

Screeds about how vile NFS is can be sent to /dev/null to save us all the
trouble of deleting them.

Jamie Bowden

-- 
"It was half way to Rivendell when the drugs began to take hold"
Hunter S Tolkien "Fear and Loathing in Barad Dur"
Iain Bowen <alaric_at_alaric.org.uk>
Received on Tue Aug 17 2004 - 15:13:54 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:06 UTC