Announcement
Collapse
No announcement yet.
Google's servers
Collapse
X
-
Re: Google's servers
Originally posted by bjaming View Posthmmm I wonder if this is some kind of silly prank...
http://perspectives.mvdirona.com/200...ncySummit.aspx
Though April 1st is a retarded day to release something like this."As Arthur C Clarke puts it, "Any sufficiently advanced technology is indistinguishable from magic". Here is my corollary: "Any sufficiently technical expert is indistinguishable from a witch"."
Comment
-
Re: Google's servers
Originally posted by bjaming View Posthmmm I wonder if this is some kind of silly prank..."\x74\x68\x65\x70\x72\x65\x7a\x39\x38";
Comment
-
Re: Google's servers
The server-in-a-box approach is interesting, Also, sticking the power supply out on the front like that is interesting, I'm guessing the power supply fan sucks the air through the whole case, which is interesting, The design is pretty neat, I'm guessing the highest quality thing in there is the power supply, then the memory... Although looking at their server-in-a-box design, its odd, it looks like their power supply is on the opposite side... Also, one of the neat things about it is that the initial design (I don't know that this is still the case) was to just leave dead items on the racks, it was more expensive to hunt them down than just ignore them, at the rate they were adding computers.I do know everything, just not all at once. This is commonly referred to as a "virtual memory" problem.
Comment
-
Re: Google's servers
Originally posted by robosycho View PostAlthough looking at their server-in-a-box design, its odd, it looks like their power supply is on the opposite side...
Also, one of the neat things about it is that the initial design (I don't know that this is still the case) was to just leave dead items on the racks, it was more expensive to hunt them down than just ignore them, at the rate they were adding computers.45 5F E1 04 22 CA 29 C4 93 3F 95 05 2B 79 2A B0
45 5F E1 04 22 CA 29 C4 93 3F 95 05 2B 79 2A B1
[ redacted ]
Comment
-
Re: Google's servers
Originally posted by bascule View PostI don't think that's the case at all anymore. All of the systems have diagnostic lights on the front (so hunting them down becomes a matter of spotting the yellow or red lights) and everything is held down with velcro for easy replacement.
I do like the idea of having the onboard battery, but I'm curious as to the charging circuit. Is it under the harddrives?A third party security audit is the IT equivalent of a colonoscopy. It's long, intrusive, very uncomfortable, and when it's done, you'll have seen things you really didn't want to see, and you'll never forget that you've had one.
Comment
-
Originally posted by robosycho View PostAlso, sticking the power supply out on the front like that is interesting, I'm guessing the power supply fan sucks the air through the whole case, which is interestingOriginally posted by bascule View PostLooking at the second picture it appears that the PSUs are in back with all the cables easily accessible in front. I see the battery (likely the #1 component that needs replacing) in the very front of each server, and it can probably be swapped without removing the system at all.
Originally posted by streaker69 View PostI do like the idea of having the onboard battery, but I'm curious as to the charging circuit. Is it under the harddrives?Thorn
"If you can't be a good example, then you'll just have to be a horrible warning." - Catherine Aird
Comment
-
Re: Google's servers
Originally posted by Thorn View Post
No, look at the enlarged picture and trace the thick red and black wires from the battery. They run right back to the power supply.A third party security audit is the IT equivalent of a colonoscopy. It's long, intrusive, very uncomfortable, and when it's done, you'll have seen things you really didn't want to see, and you'll never forget that you've had one.
Comment
-
Re: Google's servers
Originally posted by streaker69 View PostI saw that, and I was thinking that the battery is directly connected to the supply, but that wouldn't be good. You cannot continually charge those types of batteries, they just get hot and fail prematurely. There needs to be some kind of circuit in there to trickle charge them. If you'd just run 12V @ 5A into them, they'd die wouldn't they?Thorn
"If you can't be a good example, then you'll just have to be a horrible warning." - Catherine Aird
Comment
-
Re: Google's servers
Originally posted by Thorn View PostProbably, but those are special 12V only supplies according to the article. My guess would be that they have something internal that regulates the charging.Originally posted by TFAAnother illustration of Google's obsession with efficiency comes through power supply design. Power supplies convert conventional AC (alternating current--what you get from a wall socket) electricity into the DC (direct current--what you get from a battery) electricity, and typical power supplies provide computers with both 5-volt and 12-volt DC power. Google's designs supply only 12-volt power, with the necessary conversions taking place on the motherboard
Their picture of the power supply says otherwise. Input Voltage 200-240VAC 50,60Hz.
So how again are they saving power?A third party security audit is the IT equivalent of a colonoscopy. It's long, intrusive, very uncomfortable, and when it's done, you'll have seen things you really didn't want to see, and you'll never forget that you've had one.
Comment
-
Re: Google's servers
Originally posted by streaker69 View PostYou mean this?
Their picture of the power supply says otherwise. Input Voltage 200-240VAC 50,60Hz.
So how again are they saving power?Thorn
"If you can't be a good example, then you'll just have to be a horrible warning." - Catherine Aird
Comment
-
Re: Google's servers
Originally posted by streaker69 View PostYou mean this?
Their picture of the power supply says otherwise. Input Voltage 200-240VAC 50,60Hz.
So how again are they saving power?
Additionally, there are nice web page that describe how less energy is lost for long runs at 220VAC than 110VAC.
http://www.bsharp.org/physics/transmission
(See Section: " High-Voltage Transmission Lines " and equations associated with that.)
Basic Summary: Amps are current. Current is a measure of number of electrons flowing "through" a wire. Electrons have mass, and when running as current, are in motion. Resistance applies per electron. More electrons mean more loss to resistance because it is applied per electron. If you could somehow put a greater difference in potential energy in the electrons being transmitted, and transmit fewer of them to get the same power (Power = Amps x Volts , P=IV) then loss to resistance could be lower for the total power transmitted at higher voltage. So raise the voltage transmitted, and lower the amps, which lowers impact of resistance, and decreases loss to heat/light/noise.
Additional complications arise when the resistance leads to an increase in temperature in the wire, which can increase the loss due to resistance which changes in many materials with temperature.Last edited by TheCotMan; April 3, 2009, 13:34.
Comment
Comment