Labs Under Fire

DamienThorn

[H]ard|Gawd
Joined
Nov 28, 2004
Messages
1,937
So, its coming up on that time of year when the weather gets warm and faculty try to find things to do other than the research they're paid to actually perform. It's also the time of year where I get to engage in my annual war with a faculty member that leaves her AC on 24/7, consumes more fast food that I've ever seen, drives an SUV, et cetera. The computer labs (in her mind) are a serious problem for power, and so they must be off whenever not in use!

I've just generated a massive email explaining to the deans and chairs why leaving the computers on and stress-testing while running security diagnostics is incredibly important (had to write to the deans because she never talked to me about any of this, just went way the fuck over my head to complain to the big wigs). What's great is that the faculty member that complained complemented me last semester; she admitted that the labs had never run more smoothly (largely because last summer I was putting in 30hrs/wk to make sure that every single computer was running perfectly). I guess it was magic that had them running really well, and that the hard work I put in meant nothing. :rolleyes:

As it stands, I think that I'll be able to leave the labs running pretty well 24/7, but we'll see. It's possible that my production will decrease over the summer while I try and deal with this nuissance. I've been hearing back from the deans and they're really impressed with the document that I've put together, and are thinking of implementing a campus-wide performance and security blitz before the fall semester, so we'll see how things go.

 
dude a campus wide borg right on!!

heres to hopeing that she doesn't win and that your able to get everyone going now!!
 
marty9876 said:
So, your borgin the whole campus then?

It'd be awesome if I could pull that off, but it'd likely (if anything came of it) be an initiative struck up by me and then implemented according to each department's IT structure. At this point, I just want to keep my own little kingdom safe from this woman.

 
THAT would be fucking impressive.

"yea i got 200 cpu's running"

"Yea I borged a campus."
 
Well, post pics of her and we will call it even.
 
NICE...that would be awesome...and with experience in the school environment...i know EXACTLY the type of person you are talking about...good job in incorperating "security diag" in your borg plea :)
 
marty9876 said:
Well, post pics of her and we will call it even.

Considering what he said about her fast food intake I don't think I want to see it. :(
 
This is the memo that I composed for the higher ups to see (some info is fudged, though not wholly incorrect):

Below is the information regarding our power usage in the xxxxx computer lab as well as a series of reasons for why the computer labs are often left on in "low power mode" over the summer semester.

Power Draw
The power supplies that we have installed in these computers have a peak theoretical output of 350W. Realistically, they peak at about 350W but cannot maintain that level of power over an extended period of time (manufacturers are well known for using trickery in calculating maximum output by looking at the theoretical maximum amperages that can be supplied along the 3.3A, 5A, and 12A rails, each of which powers different computer components). Listed are the rough power draw for the components that are operating in the background when the computers are on, but not being used.

HDD: 5-7.2W
CPU: 84W
Memory: 20W
Motherboard component: ~20-30W

The optical drives, floppy drive, USB components, media bay components, and fans are all set to draw no power until required. The monitors in powersave mode use between 4-5W, and require their peak output when they are turned on when not in power save mode. Most modern monitors require at least 140W to initially power one; these are old enough that (based on their power inefficiency) should take around 150-170W. Holding monitors in power save mode (usually) saves power overall. While being actively used out monitors draw roughly 110-130W. The power-save modes that we have implemented have the monitors "rest" at between 4-5W until "awoken." We work to keep lights off whenever possible, reducing the amount of power that is drawn from that source.

Advantages to Maintaining Constant Low Power Draws
1) When a computer is first turned on it maxes to peak power for some time; by having the computers on and minimizing the number of times that computers are turned on the max power draw is reduced, saving power.

2) By keeping the computers in a low state of power usage the hardware's shelf-life is being prolonged; much like a lightbulb, transitors in modern computers are sensitive to being turned on from a cold state - over time this will cause degradation of the computer equipment, reducing their lifespan, increasing maintenance time, and potentially negatively impacting client satisfaction as computer component fail, leaving the computer unavailable for use.

3) Each summer, especially in the past two years, the xxxx Computer Lab Support team has worked to ensure that all computers are operating properly. This has meant subjecting the computers to a series of CPU and memory intensive "burn in" tests - these are the tests that are currently being run in the computer labs. Computers are sensitive to heat and the summer offers the team ambient temperatures that begin to near the average temperatures of the computer labs when they are filled with students in the fall and winter (though in many cases the ambient temperatures over the summer are dramatically lower than in the winter semester), allowing us to approximate the lab environment though the fall and winter semesters. During the fall and winter the Lab Support team sometimes receives calls about computers that exhibit "odd" behaviour that cannot be reliably replicated. By "stress-testing" the CPU and memory (the two parts of the computer that cause random crashes, application errors, and networking problems) the team can find and resolve most, if not all, problems in all of the computer labs so that students and instructors don't have to worry about using untested computers. Stress-testing is the primary way that the Support team can ensure that client satisfaction remains high. Using this kind of software last summer allowed the team to diagnose a series of hardware concerns and address them. Without having generated at least a month of data it is unlikely that many of the hardware problems could have been discovered, and consequently client satisfaction would have decreased as computer components failed F05 and W06 semesters.

4) Currently the stress-testing software we have installed sends packets of information to a centralized data-server once every few days, letting the team know that the computers are properly connecting to the University of Guelph network and the Internet. This is important, especially in xxx, where we are testing the onboard network cards to see if they are dropping data packets (which are uses to transmit information to the internet and back to the client computer) on a regular basis.

5) Having the computers on allows the xxx Computer Lab Support team determine where security needs to be tightened; over the past year significant changes have been made to secure theWindows operating systems against virus, worm, and malware attacks. We are unable to test new security policies during the fall and winter semester because it could degrade client satisfaction and interfere with class related activities. When problems are discovered over the summer we work to remedy them and test them in the "live" computer labs. Our efforts over the past year have led to a decrease in the number viruses and illegal programs infecting/being installed in the computer labs during the winter and fall semesters. Protecting against virus attacks reduces the likelihood of a campus-wide virus problem. Establishing and testing security policies that restrict unauthorized remote access to the computers prevent malicious users from using the computer labs to infringe upon intellectual property rights, establishing pornography servers, hosting illegal file servers, or downloading data from the internet that violates the Terms of Service with the Internet Service Provider the University of Guelph uses. Without having the computers on the Support team lacks a live environment, and is much less likely to find and resolve security problems, which may lead to decreased client satisfaction and a degradation of the classroom environment during the fall and winter semesters.

I have personally been mindful of the trade-offs between power consumption and security protocols for the past year, during which I was the Senior Computer Lab Assistant. Last summer I spent a great deal of time testing security policies, testing computer hardware, and testing the software that would be used in the fall in live environments, all while striving to minimize power consumption. At the end of the last summer every single computer that was in the labs was given a "clean bill of health" and, as a result of my intensive testing, there were fewer hardware or software related problems last fall or winter than have been recorded since I began working here five years ago. If you have any questions about what I have recorded here, I can be reached at [email protected] or at extension xxx.

 
i like how "folding" is considered "low power consumption" :) but yea...there is a lot of good BS there that board people enjoy hearing so you should be all good....good job with the big words too...u seemed to use the right amount that they get excited, and think you know what you are doing, but not confusing to them.....kudos :)
 
You could always buy a Kill-a-watt meter and get actual power draws.
 
roftranspo said:
i like how "folding" is considered "low power consumption" :) but yea...there is a lot of good BS there that board people enjoy hearing so you should be all good....good job with the big words too...u seemed to use the right amount that they get excited, and think you know what you are doing, but not confusing to them.....kudos :)

Nope that is what #4 is all about....

4) Currently the stress-testing software we have installed sends packets of information to a centralized data-server once every few days, letting the team know that the computers are properly connecting to the University of Guelph network and the Internet. This is important, especially in xxx, where we are testing the onboard network cards to see if they are dropping data packets (which are uses to transmit information to the internet and back to the client computer) on a regular basis.

Good amount of BS in there.. I like it!

 
Tormond said:
Nope that is what #4 is all about....
I know but #1-3 goes into how you are saving energy...mmmm....good old american BS at its finest...i love it :)
 
Tormond said:
Nope that is what #4 is all about....

4) Currently the stress-testing software we have installed sends packets of information to a centralized data-server once every few days, letting the team know that the computers are properly connecting to the University of Guelph network and the Internet. This is important, especially in xxx, where we are testing the onboard network cards to see if they are dropping data packets (which are uses to transmit information to the internet and back to the client computer) on a regular basis.

Good amount of BS in there.. I like it!


Learning to write semi-competently and to my audience is one of the central things that my undergrad in philosophy has taught me. Now, at the master's level, I think I'm just expected to write to an upper-tier audience and ignore the rest of the world exists :p

As for folding being low power; I don't lie about power consumption. The CPU is going to draw 84W regardless or whether or not it's folding. We don't have a step-down program on the motherboards because we're using celerons, which to my knowledge don't have any kind of power step-down program. I'm really quite serious in that we're using the minimum amount of power possible for having the computers on.

Now, the networking...while...I will see a dramatic decrease in my daily point totals if I have issues with the network, and the lab that we just replaced had chronic networking problems. I'm pretty sure that it had to do with the (shittacular) SIS900 chipset, but I really need to test the network on a relatively random basis to see if its a cabling problem. Doing it with F@H just happens to be easier that trudging through the network for hours on end :p

1) and 4) are probably the most fudged parts, though they're completely honest and I can easily defend what I've written if called on it. That said, I don't think that I will - the deans just wanted something to shut up some of the eco-freaks on campus (in addition to this faculty member a horde of students are threatening to cause problems for the labs as well. this memo might beat them before they begin their crusade.)

 
DamienThorn said:
Canadian BS, thank you :cool:

Sorry....but it sounds like good old american BS....either way...good job on BS and GREAT job with the borg...especially if you get to expand ;)
 
DamienThorn said:
1) and 4) are probably the most fudged parts, though they're completely honest and I can easily defend what I've written if called on it. That said, I don't think that I will - the deans just wanted something to shut up some of the eco-freaks on campus (in addition to this faculty member a horde of students are threatening to cause problems for the labs as well. this memo might beat them before they begin their crusade.)


What kind of problems are they threatening? Are they just concerned about the computer labs having the machines on when school is out or is that only a part of their crusade? Jerks, I say let them save their energy and let us cure cancer.
 
I know you already sent the letter, but the part about electronics "wearing out" isn't exactly accurate. Yes, you do get electron migration. But it's related to temperature, and you have to get some pretty extreme temperatures to see the effect in any reasonable amount of time. Unless you have zero airflow or a clogged HSF, you're not going to get those temperatures, and you won't see permanent effects any time while the Uni keeps the computers.

Testing for stability, though, is a good point. It's "either we find it during the summer while stress-testing, or our users find it when they're working on critical projects".

Do you get to keep folding during the school year?

Another point--processors *do* use less power when not folding. When idle, the processors go through tight, fast little do-nothing loops that require (relatively) very little power. But you don't have to say that... ;)
 
Where can we find her. I'm sure after a visit that she won't be a "problem" any longer. With the right amount of persuasion I'm sure she can be convinced to support you 100% in getting the whole campus borged.

 
roftranspo said:
Sorry....but it sounds like good old american BS....either way...good job on BS and GREAT job with the borg...especially if you get to expand ;)


In case you didnt know, Canada is in "NORTH" America, so yes, it could be considered American BS, if you want to get technical about it.
 
MrGuvernment said:
In case you didnt know, Canada is in "NORTH" America, so yes, it could be considered American BS, if you want to get technical about it.
true true...actually the whole statement is from a George Carlin called "american bullshit"...i suggest getting it....its good stuff and this letter is defiantly follows those lines :)
 
MrGuvernment said:
In case you didnt know, Canada is in "NORTH" America, so yes, it could be considered American BS, if you want to get technical about it.

American has the reference to the United States of America; citizens of the USA are called Americans. People living on Continental North America are called North Americans. Not to be picky...I just want to be accurately identified.

 
SmokeRngs said:
Where can we find her. I'm sure after a visit that she won't be a "problem" any longer. With the right amount of persuasion I'm sure she can be convinced to support you 100% in getting the whole campus borged.

yeah, but if she's as fast-food guzzling as we've heard, who'd be willing to , uh, put themselves out, for the uh, risk?

ah, calling! Mary!!!!!
 
Or just pay off all the local fast food restaurants to refuse to sell her food....
 
or if MSG or w/e gives her stomach hell, then tell the local chinese restauraunt to lay it on thick :)
 
So how can you tell from your Stanford reports which computers are not reporting in? Are you actively reading the logs to see if any are erroring. I don't get it. Sounds like you talked and walked the walk, but if anyone really probes they are going to find nothing but a electricity guzzling program. That they are footing the bill for. A lab or two of folders is a nice chunk of money every money in both electricity and in cooling costs and who ever works and studies in those labs could be uncomfortable from the heat.

I have nothing against folding, as all should know, but I do have issues with your paper and would never feel comfortable handing that to anyone. :( Also, I have issues with people not being straight forward with people when we are dealing with a Cancer Research program, that if people just happen upon the program running they could think it was a virus or trojan. That's not what we want.
 
I have a friend who goes to the University where Damien works, and one day he's talking to me and says

'I don't get it, every computer in this computer lab is running full out...' ;)



Off-topic: I haven't been around lately and I just saw Tormond posted, he's back!
 
Back
Top