[6508] in Release_7.7_team
Re: ceres/bias switching to CUPS
daemon@ATHENA.MIT.EDU (Evan Broder)
Sat Nov 14 17:37:29 2009
MIME-Version: 1.0
In-Reply-To: <alpine.DEB.2.00.0911141359480.5357@green-arrow.mit.edu>
Date: Sat, 14 Nov 2009 17:37:14 -0500
Message-ID: <2706d8dd0911141437i2ddbf1d7pc4e4d877567c7251@mail.gmail.com>
From: Evan Broder <broder@MIT.EDU>
To: "Mark W. Manley" <mmanley@mit.edu>
Cc: Geoffrey Thomas <geofft@mit.edu>, release-team@mit.edu, ops@mit.edu
Content-Type: text/plain; charset=ISO-8859-1
X-Spam-Flag: NO
X-Spam-Score: 0.00
Content-Transfer-Encoding: 8bit
I've got a changeset in -proposed now to point all cluster machines
directly at printers.mit.edu. I think that should be ready to deploy
to the clusters on Monday.
Since there are queues currently being served by cups.mit.edu that
aren't DORM or CLUSTER (we added a few on request), I'd like to have
an easy means for users to add those. We've discussed having an
"add-athena-printer" script that would do something sufficiently
clever, and it shouldn't be hard to write.
Once we have that available, I'm definitely ready to re-point
everybody else at cluster-printers.mit.edu and/or let you guys steal
the cups.mit.edu name. Hopefully we can do this within the next week
or two.
- Evan
On Sat, Nov 14, 2009 at 2:07 PM, Mark W. Manley <mmanley@mit.edu> wrote:
> I have deployed cluster-printers.mit.edu in addition to printers.mit.edu to
> help with your CUPSy needs. What's the difference, you may ask? The
> "cluster-printers" address goes to a pair of virtual servers that only have
> the DORM and CLUSTER printers from moira defined to them. The "printers"
> address, however, has 'em all.
>
> The "cluster-printers" address will help in the clusters and dorms where
> people want to have a sweet drop-list of printers, but don't want to wade
> through. I'd like, if possible, to see that used in the clusters soonish in
> lieu of the 'cups.mit.edu' address, which I would love to see folded into
> the printers or the cluster-printers name.
>
> I have also discontinued the LDAP hooks. They were creating more problems
> than solving and the CUPS support for 1.4.x LDAP was, at best, craptastic.
> My favorite was not only that the servers would just crash if they had any
> sort of blip in their connections, but that they'd also reliably
> miscalculate string length by 1, so you'd get the "cere" printer in "W2" and
> so on. LDAP would have been super-sweet rather that the clunky CUPS-style
> polling, but life is hard.
>
> Please let me know if you have any questions.
>
> -MM
>
