[6065] in Release_7.7_team
I propose we change Athena to power off video monitors.
daemon@ATHENA.MIT.EDU (William Cattey)
Tue Sep 23 17:21:26 2008
Mime-Version: 1.0 (Apple Message framework v753.1)
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Message-Id: <833CC23B-2BCD-42B9-A2D7-4E4C29D00A96@MIT.EDU>
Cc: Thomas Smith <tjsmith@MIT.EDU>, Joyce Gomes <jogomes@MIT.EDU>,
owls@MIT.EDU, release-team@MIT.EDU, "Michael R. Gettes" <gettes@MIT.EDU>
Content-Transfer-Encoding: 7bit
From: William Cattey <wdc@MIT.EDU>
Date: Tue, 23 Sep 2008 17:20:30 -0400
To: Chris Lavallee <lavallch@MIT.EDU>, Chuck King <chuckk@MIT.EDU>,
John T Guy <jtguy@MIT.EDU>
X-Spam-Flag: NO
X-Spam-Score: 0.00
Summary:
Chris, John:
Presuming that we are agreed to refine the existing twice-daily
reporting and to get Hotline going with athinfo for remote diagnosing
of Athena systems, how about IAP of this year for switching over to
power-saving monitors Athena-wide?
Detail:
The question keeps coming up, "Why does Athena keep its monitors
powered up? Wouldn't it be significant energy savings?"
The Athena Release team did a back of the envelope calculation and
looked up LCD monitor power consumption. Assuming 1000 machines, 10
hours of idle time per machine, and 25 w saved by powering them off,
that adds up to about 100 MegaWattHours per year. Actual monitors in
the field actually consume more than 25 watts, but the ones we've
started buying consume about 20 watts.
The normal way the X server runs with Linux powers off video monitors
when they've been idle for about half an hour. Athena hacked the
configuration to disable that. We disabled it for two reasons:
1. We were afraid users would think the computer was broken. This
is no longer
an issue. People are used to power-saving monitors.
2. It was easy to stand at the doorway of an Athena General Use
Cluster, and
determine at a glance which systems were offline. This is the
remaining reason
why we keep the monitors powered up.
I discussed this situation with John Guy. He expressed the concern
that, as we continue
doing more work with fewer resources, an easy way to identify sick/
healthy systems
is important.
The Athena release team discussed the principle of making it easy to
identify sick and
healthy systems that still allows powering off of monitors. There is
a twice-daily report that provides a variety status information (up,
down, hacked, hung, etc.) status on the Athena systems we take care of.
There is also the athinfo program that enables remote query of Athena
systems for additional information. I confirmed with John Guy that
the twice-daily report is used by Hotline to identify troubled
systems. I have recently learned that the Hotline team hasn't been
trained in making good use of athinfo.
Bottom Line proposal:
1. We agree to lean more heavily on the twice-daily report, and to
refine it so it really meets the needs of Hotline.
2. We train Hotline in use of athinfo, and refine it if necessary to
offer additional queries of importance to most effectively remote
diagnosing systems.
3. We set a date for removing the hack to keep Athena monitors
powered up when idle.
I believe John Guy is the primary stakeholder and his opinion going
forward should be most heavily weighted.
I believe Chris Lavallee is the decision maker here.
Chris, John: How about IAP of this year for switching over to power-
saving monitors
Athena-wide?
-Bill
----
Important: IS&T IT staff will *NEVER* ask you for your password, nor
will MIT send you email requesting your password information. Please
continue to ignore any email messages that claim to require you to
provide such information.
----
William Cattey
Linux Platform Coordinator
MIT Information Services & Technology
N42-040M, 617-253-0140, wdc@mit.edu
http://web.mit.edu/wdc/www/