🚀 go-pugleaf

First of all, let us try to avoid another flame war.

In article <459po3$pgh@Venus.mcs.com>, karl@MCS.COM (Karl Denninger) wrote:
> The problem is the PROPER *MX* host for UUCP customers of MCSNet has ALWAYS
> been either "uucphost.mcs.com" (historical) or "uucphost.mcs.net".  NO
> OTHERS are guaranteed to work.

I am not a UUCP user.  To refresh your memory, you recieve all inbound
mail for solgrp.chi.il.us and route it to a mailfile.  I collect it from
there by POP3/Eudora.

I do have an MX record that was reconfigured in the fall of the original
Clout project by *YOU* and Doug Blair (who I remember as being on the
Clout side) to point to your services.  (This is when I became your
customer.)  

> Mercury USED TO (over a year ago) handle UUCP.  It then moved to Mailbox,
> sometime last year, which also handles user mail.  About a week ago (in 
> conjunction with our other UUCP changes) it was put on a dedicated 
> multiprocessor 133Mhz system to meet our promise of greatly improved 
> UUCP performance and reliability.  

And without notice or checking for what legacy uses may be in place,
Mercury stopped functioning on October 2nd.  I have had a week's outage of
email since then.

> That promise has been kept.

Gee, thanks.

> The new machine is physically called "uucp1", and the CNAME pointers were 
> changed internally to point UUCPHOST there.  90+% of our customers were
> correctly configured (including those using UUCP over TCP/IP) and
> experienced NO problems of this kind in the switch.  In fact, except for 
> an "unavailable" period while disks were copied, those customers saw *no*
> changes other than performance-related ones.
> 
> ISPs have *documented* names for machines for a REASON.  Using hard-coded
> names you think do multiple things is a really bad idea, especially for
> automatons such as mail forwarding (which doesn't know enough to "see" that
> something is having a bad effect).

Gerry Swetsky was good enough to be immediately respond to my request and
make the changes to dns.chi.il.us, and restart the name server.

However Karl, I sent you email four hours ago after Gerry went in to
request that you restart your named process so that I, your customer,
could start receiving email.  You've had time to document MCS's viewpoint,
could you please restart your named so your customer can recieve service. 
To avoid contrary claims, output follows:

> bash$ date
> Sun Oct  8 22:34:33 CDT 1995
> bash$ nslookup
> Default Server:  Kitten.mcs.com
> Address:  192.160.127.90
> 
> > set type=mx
> > solgrp.chi.il.us
> Server:  Kitten.mcs.com
> Address:  192.160.127.90
> 
> Non-authoritative answer:
> solgrp.chi.il.us        preference = 10, mail exchanger = mercury.mcs.com
 

I also requested if possible the output of a grep of the error logs I HOPE
that you keep for the amil that you bounce.  This is so I can attempt to
pickup whatever business corresponces that have been dropped.

I would appreciate the DNS cache flushed ASAP, and a list of the bounced
email by tomorrow.  Thanks.

-- 


   Walt Pesch                   Systems Solution Group
   312/315-4077                 walt@solgrp.chi.il.us