From: "James W DeRoest" <deroest@cac.washington.edu> To: "AST Webpage" <astweb@u.washington.edu> Subject: Light bulb for downtime Date: Mon, 22 Dec 1997 07:56:29 -0800
-----Original Message----- From: Thomas W Remmers <remmers@u.washington.edu> To: Eliot Lim <eliot@u.washington.edu> Cc: David Wall <davidw@u.washington.edu>; Yonah Karp <yonah@u.washington.edu>; Oren Sreebny <oren@u.washington.edu>; Douglas Luft <dbluft@u.washington.edu>; Jim DeRoest <deroest@u.washington.edu>; hubert@cac.washington.edu <hubert@cac.washington.edu>; Jim Fox <fox@u.washington.edu>; corey@cac.washington.edu <corey@cac.washington.edu> Date: Friday, December 19, 1997 10:33 AM Subject: Re: tentative downtime schedule
>I currently have queen.u.washington.edu configured as a lightbulb for
>the downtime on December 28. Any telnet, rlogin or rsh to
>queen.u.washington.edu, followed by entering a valid UA username only,
>will cause a message to be displayed:
>
>------ cut -------
>
>The University of Washington Uniform Access computer you are trying to
>contact is currently unavailable due to a previously scheduled and
>announced period of systems maintenance. The computers Homer, Dante,
>Mead, Saul, Goodall and Aagaard are scheduled to return to service 4AM
>monday morning, but may be available before that time.
>
>Computing and Communications regrets any inconvenience, but systems
>maintenance is necessary for the stability, integrity and performance of
>the computers.
>
>Other services that may be affected are the library systems, ftp files
>residing on ftp.u.washington.edu, Web pages on weber.u.washington.edu,
>and UWIN. Email sent to Uniform Access accounts will still be
>delivered.
>
>------- end ----------
>
>Please review this text and how it appears to users. Something similar
>can also be used for the Dante downtime this Sunday. During the
>downtime, the mead, homer, saul and goodall name service will point to
>queen. Users who connect directly to a particular host will not receive
>any message, but little can be done without modifying DNS.
>
>Tom
>
>On Wed, 17 Dec 1997, Eliot Lim wrote:
>
>>
>> (for those just tuning in, we're planning a big upgrade of all the AIX
>> boxes during the qtr break.)
>>
>> > 1) will modem pool authentication still work?
>>
>> yes
>>
>> > 2) what does 'library systems' mean?
>>
>> haley, hopper, libsrv. i guess that would affect uwin wouldn't it. tom?
>>
>> > 3) will UWIN still work?
>>
>> uwin machines will be affected, which would affect uwin clients
everywhere.
>>
>> > 4) weber (serving of pages)
>>
>> weber itself is a DEC box which we will not be patching this time around.
>> (only AIX boxes really) however, people's web pages are NFS mounted on
>> AIX boxes, so weber will respond to queries but will not be able to
>> deliver the pages. i guess we should kill httpd on them. i'll
>> let jim fox comment on this one.
>>
>> > 5) ftp.u
>>
>> same basic answer as above. ftp.u is a DEC box, but the files reside
>> on an AIX box.
>>
>> > Are we correcting in assuming the only interruption in email message
>> > sending and delivery would be some queuing on MX systems while a
>> > particular inbox server is un available?
>>
>> i think so.
>>
>> > I.e., the jasons won't be out of service?
>>
>> i have already upgraded a couple of jasons, so even if i haven't done
them
>> all by then there will be at least a couple still standing since they've
>> already been upgraded and rebooted.
>>
>> > The mailhost.u servers will not be offline? etc.
>>
>> are these NDC boxes?
>>
>> > li won't be down?
>>
>> li will be affected but we will stagger the outage so that the backup
>> server can respond while the primary is down. will talk to jim fox
>> about this.
>>
>> > Anything we can anticipate that would affect help@cac email, or the QnA
>> > system?
>>
>> don't think those would be affected. i'm CC:'ing steve hubert on this.
>>
>> i think bob jamieson also wants some notification given to the batch
>> machine users. we'll work something out.
>>
>>
>> eliot
>>
>>
>
>