[2937] in SIPB_Linux_Development
Re: Hesiod information for systems installed with our installer
daemon@ATHENA.MIT.EDU (Greg Hudson)
Mon Aug 28 16:05:38 2000
To: Sam Hartman <hartmans@MIT.EDU>
Cc: linux-dev@MIT.EDU
In-Reply-To: Your message of "Mon, 28 Aug 2000 14:53:35 EDT."
<200008281853.OAA27620@sweet-transvestite.mit.edu>
Date: Mon, 28 Aug 2000 16:05:15 -0400
From: Greg Hudson <ghudson@MIT.EDU>
> I had assumed that we could install local cluster information for
> machines without Hesiod, but this is a bad idea.
Couple of things you didn't appear to consider:
* We (Athena) kind of like to have a central list of machines
we might screw over by making various changes to the
release. We might be able to maintain such a list in other
ways, but Hesiod has some nice advantages. Unless people
start using cluster.local a lot.
* The Hesiod method gives us a lot of flexibility to make
update policies on the fly. If Hesiod information contained
no version information, and the machine instead had local
state saying whether it takes alpha or beta releases or
whatnot, then we couldn't auto-update arbitrary groups of
machines on different nights, for instance.
