Path: visi.com!winternet.com!not-for-mail
From: mfrankow@winternet.com (Michael Frankowski)
Newsgroups: winternet.announce
Subject: The Week
Date: 12 Jul 1996 22:33:52 GMT
Organization: StarNet Communications, Inc
Lines: 75
Approved: mfrankow@winternet.com
Distribution: local
Message-ID: <4s6jsg$gvr@blackice.winternet.com>
NNTP-Posting-Host: tundra.winternet.com
Summary: The week's happenings.
X-Newsreader: TIN [UNIX 1.3 950824BETA PL0]

Winternet Users,

Yesterday we had more interesting happenings at Winternet. Due to the
long processing time that our sendmail server takes to verify that a
user does not exist it shut itself down at about 2 a.m. and our
backup servers took up the load for us, turning around and driving our
load up again at 9:30 a.m.  when we brought the service back up.We
spent yesterday riding out a 'mail-storm' as our two backups kept
pounding us with messages that could not get threw the first time.  A
thank you goes to Daniel Frankowski for keeping the mail machine
running as much as possible.  It is estimated, by our backup servers,
 that we lost about 2,500 pieces of mail during this time.

During this outage the two machines that take in mail started putting
it in two different places so it looked like people would send
winternet a letter and it would step off into the void.  You couldn't
read those letters. This has been solved and all the "lost" letters
have been added back into your mail box.  It appears that this has
happened before so a few users might have received some very old
"missing" letters.

The phone system - We ARE here!!! But for part of the time the voice
mail system would pickup the phone in a half-ring. so we never had a
chance to pick it up.  We have had the phone repair people out to over
haul the system, it should now be working correctly. We also thank you
for the phone calls to tell us what you cannot do, that is one way we
have for finding out what is going on 'Out There'

Our technical support does get a list of calls, as does everyone
else. We are going down our lists as fast as we can to give you a call
back if you need help.

Yesterday also saw us lose access to the personal home pages as
www.winternet.com had a process table corruption. The machine was
restarted and came right back up and went back to work.

For a ten minute period around 11:30 am our local network flooded,
causing yet again the infamous "password rejected" problem, and it slowed
everything else down as well. The router performed a hard reset of the
ethernet network and cleared up the problem automatically.

And I made a mistake yesterday. At about 10:30 am I commented out a
few lines in the Domain Name System (DNS) database using a '#' instead
of a ';'. This caused a few strange problems relating to name
service. I corrected the problem at around 1:30 pm. I apologize. I
have to remember that I am not programming shell scripts all the time.

I would like to thank Jim (JR) Williams for coming down and helping
with the E-Mail problem, and all the other people who offered advice
on the problem over the phone. It is nice to have friends.

As many of the more technically inclined have noticed at this time we
are not using our AlterNet link in an even balance with our two MR-Net
Links. This is because we would rather keep your access to the net
available so the routing table in the large CISCO is still not
perfect. Over the next week we will be balancing out load across all
three links again.
 
Over the last week, Winternet has faced some of the toughest
challenges we've ever encountered.  It has been a difficult and trying
time for all, most of all for our customers.  We would like to 
sincerely apologize to each and every Winternet customer who feels that
they have not received the quality of service they deserve.  We hope
that this message and the ones to follow will demonstrate that we are
strongly committed to communicating with, and serving each of you. 

Thank you for your patience and support.

More news to follow....