Slashdot Mirror
Why IT Won't Power Down PCs
snydeq writes "Internal politics and poor leadership on sustainable IT strategies are among the top reasons preventing organizations from practicing proper PC power management — to the tune of $2.8 billion wasted per year powering unused PCs. According to a recent survey, 42 percent of IT shops do not manage PC energy consumption simply because no one in the organization has been made responsible for doing so — this despite greater awareness of IT power-saving myths, and PC power myths in particular. Worse, 22 percent of IT admins surveyed said that savings from PC power management 'flow to another department's budget.' In other words, resources spent by IT vs. the permanent energy crisis appear to result in little payback for IT."
I worked as head of Critical Factilities Engineering for a major financial services provider with a 1 MM sq ft campus. There were just over 4000 employees on the campus, each one with at least 1 computer at his/her office/cube. After having a very expensive energy audit performed, a potential savings was (big surprise) shutting down PCs.
Despite calculating that the organization could save $75K annually (this was a conservative estimate), their marketing department put a stop to the idea. Why marketing? Because the company had just gone through a "rebranding" and the marketing department had designed a new screensaver for all workstations with the new logo/slogan. None of these computers were in client facing positions, so effectively, they were insistent on wasting energy to advertise....to themselves!
No, I'm not kidding.
99% of the time, if I'm not sitting in front of it reading Slashdot, my work PC is merrily chugging along folding proteins and using up company electricity.
But that other 1% of the time, I'm using it from home, because I've gotten called up to fix some urgent client problem.
To save that $75 worth of electricity, my company would have to require that I drive in to the office every time a client has a hiccup that I can diagnose and fix in five minutes. I don't get paid by the hour, but I'm fortunate enough to work someplace that values my time -- including my non-work time. They would consider that $75 to be money well spent to keep me able, and most importantly *willing*, to take time out on a Saturday to fix a simple problem.
Stressed? Me? Of course not. Stress is what a rubber band feels before it breaks, silly.
It's a pain in the ass, no one really cares, and the first time some manager had data loss from a machine shutting itself down, the policy would end.
If we all sat down and set up our networks so that everything correctly booted and shutdown when the network told it to, we could attach power management stuff to the whole network...Assuming that everything correctly saved state when it shut down, so that people didn't lose all their work when their machine automatically shut itself off.
They're treating this like it's just lazy admins, but its a knotty problem, and not a particularly critical one. In datacenters the computers are the primary energy draw, in office buildings it's light and climate control, and, judging by the heating bills in the winter, the computers aren't really heating the building up that much.
ad logicam Claiming a proposition is false because it was presented as the conclusion of a fallacious argument.
There are a number of admins out there that won't power down any server even if it's the only way to fix a problem that's trashing their files and network.
Sometimes it's because they don't 'have the authority' to down the machines.
Other times it's because they get unrealistic bonuses for unbroken uptime, and they are greed cretins who'd rather see their work go down the tubes for money.
I know that it's rarely an issue with downing non-servers, and most admins are responsible as well as being the rarely disputed managers of their boxes, but there are way too many fools and scum.
If you're curious, yes, I've dealt with a large number of those two types I just listed. They have no pride in their work, and give all admins a bad name. But that's all fodder for a different rant.
The OS and hardware should incorporate power saving into machines that are logged out.
Our users are instructed to logout, but to leave their machines on for patches and the like.
If the OS could detect when the user was logged out and no services in the background where doing things we could
really turn down the machine.
A logged off machine's cpu could virtually go to sleep, the harddrives slow to 5200rpm or lower, the monitor go to sleep, and so on.
yes it's not as good as shutting the computer completely off, but maybe with some better types of wake on lan we can get as close as possible. Or scheduled turn on and off. Like tell windows to shut off from 7:00 P.M. till 1:00....turn on to get updates and then shut back down.
Ulitmately this just needs to be the default for future version of OSs like windows and the like. I think we really have to make it a brain dead for IT as possible. I've got enough other crap to worry about...although I do worry about the world engergy problems.
Nothing built into XP, Vista, or Group Policy supports time-of-day power management. Many cases the user never wants their PC to sleep/hibernate from 9-5, but after 7 it's fair game. Microsoft doesn't address such a situation. It's either all-or-nothing. The alternative is to spend a lot of time/money acquiring some 3rd party tool like Verdiem, but buying an enterprise tool, versus enabling a feature you already have, means most people won't do it.
I teach physics at a community college, and I recently made a big push to get proper power management set up in the science division's computer labs. It ended up being orders of magnitude more work than I thought it would.
I had seemed like a total no-brainer to me. We had 42 desktop Windows machines in our student computer labs. They were running 24/7. They had CRT monitors, and they were configured so that when they weren't being used, they ran a waving flag animation on the screen, meaning that both the CPU and the monitor were drawing full power. Here we were teaching our students about global warming, but we had this ridiculously wasteful configuration.
The first issue was that, as the slashdot summary suggests is common, nobody really cared, because it was some other part of the organization that was paying the electric bills.
The second issue was that when I approached IT, they wanted to handle it using software called Deep Freeze, which not only handles power management but also automatically restores the computer's hard disk to a known state every so often. This is in principle a good idea, because it means that students can't screw up the machines, and it's another layer of defense against malware. However, it opened up a whole can of worms, because if they were going to make this new hard disk image, they wanted to make sure it was done right. They wanted to update the OS, and install all the apps from scratch. Well, we had a ton of apps dating back to ca. 1995 that were still being used for instruction, but nobody could find the licenses for them. So that became a huge issue. It was one that we would have had to deal with sooner or later anyway, but it was a clear example where the easiest thing to do is always to leave things the way they are.
So we finally got that done, after much interpersonal conflict and hurt feelings. Now we have the new issue, which seems to be that Deep Freeze doesn't play nicely with Windows updates. In one lab, for example, we have about 60 machines, roughly half belonging to the science division. Their hard disks get reimaged over the weekend by Deep Freeze. But wait, then on Monday morning people walk into the lab and power up all the machines. Now all 60 machines phone home and realize that they need an update from MS; they had the update before, but it got erased by the reimaging. So they all start downloading the same 100 Mb update at once, with predictable effects. A chemistry teacher brings in a whole class to do work on the computers, and the computers are completely unusable. Oops, time to come up with a new lesson plan. Hope he's good at thinking on his feet.
Of course there's no reason in principle that all of these different issues had to be coupled together. E.g., Faronics, which sells Deep Freeze, has another product that only does power management, not reimaging. But the thing is, in real life you're dealing with complex systems and complex human organizations, and lots of well-intentioned changes can have unintended effects.
Find free books.
I implemented a nightly shut down policy for our users because I got sick and tired of them lying to me about the last time they rebooted their PC.
"Everything is running like crap"
"Have you rebooted?"
"Yeah, like 5 times."
*walk over to PC, bring up command prompt*
-net statistics server
"Statistics since 8:00AM at ."
*facepalm*
I pitched it to management as power savings, but really I could care less. I just wanted to have a way to force those bastards to reboot every night. And yes, it did make a pretty significant difference in the amount of support calls I got. I suppose you can thank Windows XP for saving power, haha.
PS-Is it wrong for a sysadmin to hate his user base? Even if they're really, really stupid, because your company is cheap and only hires incompetent morons (excluding the sysadmin, naturally...)?
As our way of thanking you for your positive contributions to Slashdot, you are eligible to disable Slashdot 2.0.
... then we'd power them off. But, undoubtably this would lead to helpdesk calls.
I have responded to "dead PC" calls when, in fact, the PC was not plugged in, monitor not turned on, etc. At one job, that was like 20% of the work load.
Give a man a fish and you have fed him for today. Teach a man to fish, and he'll say "WHERE'S MY FISH, YOU IDIOT?"
You can't send wake on LAN packets across the Internet because the Internet is not a LAN. WOL packets are Ethernet frames, not IP packets. They will not be routed because nothing high enough up the protocol stack to handle routing ever sees them. That said, there's nothing stopping you from having a machine in each LAN segment that is always on and provides a web UI for starting machines by sending a WOL packet.
I am TheRaven on Soylent News
Our group was recently informed "the simple act of shutting down PCs at night can save a company with 10,000 PCs over $260,000 a year". We kicked around the idea.
That's an alleged savings of $26/year/computer, or about $0.09/day.
Assuming it takes 10 minutes daily to turn a computer on, wait for boot, and fiddle with getting everything back up to where it was*, we're looking at something vaguely around $6.00 spent just to recover from "the simple act of shutting down [an employee's] PC at night".
So turning off the computer at night costs roughly 64 times as much as leaving the durn thing on.
(* - I've got 20 windows open right now, and half of them took considerable time to get to where they are now as I'm debugging something.)
Can we get a "-1 Wrong" moderation option?
In many companies, you would be darn near shot for asking if you can access your PC from home. Besides, what do you need off your PC? Your data should not be on the pc, it should be sitting on a server, where it can be properly secured. That's what VPN's and terminal servers are for..
Which is fine. My position is simply that IT shouldn't whine about atypical user behavior patterns unless/until they are ready to address their underlying business needs . . . which, frankly, IT is often ignorant of or indifferent towards.
Updates should NOT be run at night, because then the machines never get rebooted to actually activate the updates (unless you tell them to reboot even if the user is logged in, which ruins the "running a simulation all night" thing)
Yeah, I love policies like this. I'll be setting in some seminar and in the middle of the speaker's presentation there will be a forced reboot after a background install. Lovely. Convenient for IT, yes. Inconvenient for the audience of 100+ who get to wait through a 5 minute restart. Or, hah, I'll turn on my PC at the beginning of the day and run through the ungodly boot times imposed by who knows how much crap that's been layered on by IT, followed by the obligatory virus scan, installs, reboots etc. Some days, it's 30 minutes to an hour before I can get to work. Of course IT doesn't mind about stuff like this because opportunity cost is unmeasured and in any case doesn't hit their books. For all the processes I had to deal with in a large corporate environment, I never saw a full requirements analysis run before IT rolled out a policy, or an impact analysis afterwords. Lesson learned: IT cares about IT and not much else.
But life is good. I'm in a smaller company now, the IT people actually care if projects and work are getting done, and the users try not to make life miserable for IT (and sometimes succeed) -- so much more comfortable than having to fight IT day by day just to do the work I was hired to do.