[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[cobalt-users] Questions on Cobalt from a linux user...
- Subject: [cobalt-users] Questions on Cobalt from a linux user...
- From: Charlie Summers <charlie@xxxxxxxxxx>
- Date: Sat Jul 14 05:12:18 2001
- List-id: Mailing list for users to share thoughts on Cobalt products. <cobalt-users.list.cobalt.com>
Folks;
First, much thanks to all of you who have answered questions for me
without knowing it...I've been spending a good bit of time in the archives in
the last week, and the assistance has been invaluable (although the search
engine leaves some to be desired...searching on a specific version number
fails gloriously).
I'm coming to the Raq3 after being comfortable (not a "guru," mind you,
but comfortable) with the administration of "real" linux servers, so I'm
unfortunately not a "newbie" and as such find the web interface to be a
hinderance rather than a help. Everything on the machine seems to be either
in the wrong place, or an outdated version (even with all the patches - hey,
at least there's a reasonably modern version of BIND, so I shouldn't gripe).
But I've been trying to use the interface to do the bulk of the work even
though it's slow, awkward, and I've grown to hate it, because I don't really
know what the results would be if I _didn't._
But I want to install SmartList (mailing list software based on procmail),
and am used to creating a user "slist" to manage the lists. If I skip the
interface, and just create the d*mned user like a real administrator, will
the sky fall and the machine blow up the way Sun (and my provider) seem to
imply that it will? I don't need a user attached to some virtual web site, I
just want a regular old-fashioned user - is this ok to do, or will Sun void
my provider's warrantee and take my first-born child? (My provider was
stunned that I could manually create the eth0:0 network-script manually in
just a few seconds to attach a second IP to the machine...apparently you're
supposed to spend ten minutes in the interface and create a dummy virtual
site pointing to that IP; making a ~100-byte text file seems to be a major
no-no. ;)
Also, while I'm asking silly questions, I was planning on using the POP
Before SMTP Relaying package, but have been told that there is a major flaw
in the routines that allow spammers to inject their IPs directly into the
thing permitting RELAYing - is this true, or does the person who told me this
not know whereof he speaks? (The routines I use now can't possibly work on
the RaQ, since they directly rewrite /etc/mail/access with the "new"
POP-authenticated relays at the top of the access file every couple minutes,
adding the contents of a separate file holding the user-maintained data to
the bottom, which would "eat" changes made by the interface, if I understand
the way the RaQ maintains the access file.) If this _is_ true, anyone have
any thoughts on how to close the security hole? I'd rather force my users and
myself to use a local provider's SMTP server than take the chance the scum
will use MY server to send their garbage, you know?
If I sound like I don't like the RaQ, I really do, I'm just a little
frustrated by all of the seemingly unnecessary changes between a stock linux
system and the system on the RaQ. It should be child's play for an
experienced admin to handle, and I'm finding it...less so.
Charlie (who gets to install PHP today. Yippee.)