How Many Desktop PCs Can One Server Replace? 107
NZheretic asks: "HP has just announced that they have upgraded a four-processor server with Advanced Micro Devices' new dual-core Opteron. The amount of processing power a multi-processor multi-core system can deliver seem like a waste of processing power for most traditional servers, which are more likely to suffer from disk access bottlenecks before lack of processing power becomes a problem. But what if that power could be delivered direct to the desktop users? The HP ProLiant DL585 supports eight 64-bit PCI-X I/O Slots (Six 100MHz, two 133MHz). The ATI FireMV(TM) 2400 supports Quad DVI/VGA displays on PCI Express. Assuming that you leave one PCI-X slot for a multiport USB card, thats up to twenty eight displays with USB keyboards,mice and headsets that could theoretically replace twenty eight networked desktop PCs. Using DVI and USB extenders, not all of the user stations would have to be within the 7.5 meter cable distance imposed by the DVI cable limit. The only OS currently capable of supporting this many displays is Linux. What limits would be imposed by the hardware and PCI-X bottlenecks? Taking into account the added cost of the HP and ATI hardware, could it deliver a great reduction in the total cost of ownership over both traditional PCs and thin client systems? How many desktops is it practical for a high end server to directly replace?"
One Giant Step Backward (Score:5, Insightful)
I'm sorry, but this is one of the dumber products I've seen out there.
The software retails for $99 per workstation, and this gets you only one year...additional years are $29, again per station.
Add to that cost the cost of all those dual-headed video cards, USB cards and hubs, and DVI and USB extenders, and your total cost is not at all inconsequential. And for all this work, you have a maximum of 10 users to a server? Plus, those users are physically tethered to the server, severely restricting your network design.
It seems to me that all this and a lot more could be accomplished with less money and less hassle via some very low-end systems and VNC. In fact, that's how I'm accomplishing it right now.
Re:One Giant Step Backward (Score:2, Interesting)
For a regular setup, assume a 3-year upgrade cycle and a $1200 computer (not unreasonable).
As long as you can do it for under $12k, you're going to come out ahead. I don't know enough about the equipment to price it all out, but I don't think it sounds that unreasonable.
Re:One Giant Step Backward (Score:2)
$1200??? I said low-end!
If you think that's a good price for low-end, I've got some Celeron 300s laying around here you might be interested in...
Re:One Giant Step Backward (Score:1)
Re:One Giant Step Backward (Score:2)
OK...that makes more sense...kinda defeats the whole idea of the thin-client/dumbterm solution, but that raises an important question...is the thin-client/dumbterm model even viable anymore in today's era of low-priced desktop systems?
Re:One Giant Step Backward (Score:2)
OK...that makes more sense...kinda defeats the whole idea of the thin-client/dumbterm solution, but that raises an important question...is the thin-client/dumbterm model even viable anymore in today's era of low-priced desktop systems?
Last I saw, you could pick up a new box (no display) for 250 dollars. That included Windows XP.
That's retail at CompUSA.
Re:One Giant Step Backward (Score:2)
"TCO > Cost of hardware and software."
The thin client model mainly works because, properly done, the users cannot break their PC. Even if they throw it out the window, or set fire to it, they cannot break it. (If they do that you get another thin client out of the cupboard and plug it in, presto, back online in 5 minutes)
However, the thin client model does mean that the direct attach model has a hard time
Also, there are other reasons for using thin client. For example, I have several
Re:One Giant Step Backward (Score:5, Interesting)
It gives the same end result without messing with exotic hardware and configurations, and you only have to be as close as your nearest ethernet port.
Re:One Giant Step Backward (Score:2)
I've been considering an option like that for the bedroom for a while now, but there's a morass of info to wade thru. Couple $150 with the cost of a cheap 15" LCD and it's in my price range
SB
Re:One Giant Step Backward (Score:2)
NTAVO [ntavo.com]
I've used it with TS and X, and it works pretty good. The fit and finish on the hardware is excellent. It boots linux and has an interface that is really familiar to Windows users. (start button, status bar, icons on the desktop for various server connections)
I bought the first one to demo to clients, but I think I'll get a couple LCDs and put one in the living room and another in the kitchen.
Re:One Giant Step Backward (Score:1)
My one gripe is that the LCD model would be better replaced with a notebook at that kind of pricing. Still damn sexy though.
Re:One Giant Step Backward (Score:2)
Disagree (Score:2)
Re:One Giant Step Backward (Score:2)
Works OK. $2000 box. Snapshots and backups make it worthwhile.
So a 4x2-way HP with Ultra320 SCSI drives ought to be able to handle a dozen users easily.
Re:One Giant Step Backward (Score:2)
Have you tried using Remote Desktop instead ? It's probably a *lot* faster.
Re:One Giant Step Backward (Score:2)
Are you familiar with UltraVNC? It works as a device driver which helps with the latency tremendously. It's different than regular VNC.
Over a modem I've been pretty happy with Remote Desktop, but this is over a switched 100 full network and I don't see any downside in this case to using the open standard, always my first choice.
Re:One Giant Step Backward (Score:2)
mount local hard drives to the 'serving' machine automatically?
load local printers to the apps running on the serving machine automatically, then remove them when disconnected?
Not tie up the serving pc from being used locally?
My work, I'm the IT guy in addition to a lot of other stuff,
the web is LOCKED DOWN at work via a 3com router with a short 'allowed' url list.
however, I can RDC into my XPPRO 'chine at home, and do all my webbrowsing,email/whatever- while my wife plays DiabloII on t
Re:One Giant Step Backward (Score:2)
Well, when you're running several virtual machines on a single server there's not much point to local access, now is there? Terminal Services is good for certain problems but that doesn't mean its the right choice for every problem.
But yes, someone at the console could interact with the remote users, should it be needed.
Re:One Giant Step Backward (Score:2)
But VNC is a better idea. We do provide ~20 users with terminal services desktop on a windows2000 server (dual PIII 1.4GHz), and it has been very snappy and impressive for 3 odd years now. No plans to upgrade it. At one time we had ~20 Pentium1 workstations using the server; savings and easy administration.
The only issue appears with games and movies. Movi
Re:One Giant Step Backward (Score:2)
Re:One Giant Step Backward (Score:2)
Section "InputDevice"
Identifier "Keyboard0"
Driver "evdev"
Option "Device" "/dev/input/event0"
Option "AutoRepeat" "500 50"
Option "XkbLayout" "us"
EndSection
Section "InputDevice"
Identifier "Keyboard1"
Driver "evdev"
Option "Device" "/dev/input/event1"
Re:One Giant Step Backward (Score:2)
Re:One Giant Step Backward (Score:2)
If it were me.. (Score:3, Funny)
-m
can't you do this anyway with native X? (Score:2)
The only issue would be the USB/user thing... but with hald/gnome-volume-manager (not sure of the KDE equiv), this can be worked around...
Why buy this product when I _THINK_ you can do this already.
ltsp (Score:3, Insightful)
What's old is new again! (Score:1)
Re:What's old is new again! (Score:2)
The economic problem with these approaches is that you'd have to sell millions of devices before you'd get the economy of scale that PCs enjoy.
You end up with compromised functionality at about the same cost per user.
PCI-X != PCI-Express (Score:4, Informative)
Depending on how these systems are configured, it may not be possible to use that many monitors.
But this might work... (Score:2)
However, Matrox has made 4 and 8 port PCI video cards for many years. Here's is one such card based on the G200 chipset (on par with the NVidia TNT2 performance).G200 MMS on eBay [ebay.com]
Imagine... (Score:1)
Re:Imagine... (Score:2)
Since this is the anti-beowulf solution (it's one PC pretending to be lots of PCs, while beowulf is a lot of PCs pretending to be one), then if you clustered them, they would cancel themselves out in a flash of gamma radiation.
You should try it, a bit of hard radiation might do your spelling some good.
PCI-X != PCI Express (Score:1)
Your plan will not work with this motherboard.
Linux is no good (Score:2, Insightful)
Re:Linux is no good (Score:2)
Re:Linux is no good (Score:2)
While I like the advances in bus technology, there have gotten to be too many incompatible buses now. It's horrible. Looking at the ASUS web site, we now have motherboards with 5 slots, but all of them different! That's just insane. Why can't they come up with ONE bus that works at high speeds (64 bit), and works for both video and other devices? Is it totally impossible to
Why isn't Linux any good? It's been done already. (Score:2)
Re:Linux is no good (Score:4, Informative)
sure it is (Score:2)
Of course, that makes it no less of a stupid idea to do that (you should be using an X terminal and set the thing up as a server). But, in principle, Linux will support this sort of insanity if you must.
Re:Linux is no good (Score:2)
-
+
Phew. That's some hard-core hacking required to make it work. Of course, to make keyboards work, something similar will need to be done with
God that's dumb. (Score:4, Interesting)
Now buy 30 thin clients. Each one gets a KVM and a network card. Good. Now plug in the power on all the thin clients and plug their network cables into a switch. To remove clutter if you want you can use 802.11 and all the thin clients will only need power.
Ta-da! Welcome to intelligence.
Re:God that's dumb. (Score:2)
We threw away a bunch of X terminals back in the late '90s for exactly that reason. Kinda scary to see it's being suggested in 2005.
Re:God that's dumb. (Score:2)
Heck, half these things run a cut down Linux as the thin client's local OS.
Re:God that's dumb. (Score:2)
Of course, this is how I use windows now. It's just a cheap desktop that provides a web browser and an X server for my real work (all done on unix).
WAY too expensive. (Score:4, Informative)
Now add to that cost, the single point of failure issue. Even if the hardware never fails, all you need is for some malicious or clueless user to run
Addendum (Score:3, Informative)
When you do it this way, the cost goes down in dramatic fashion. A $50,000 server setup is only $250 per user when you have 200 users running off it and A server as large as you suggest could easily run 500 or mor
Re:WAY too expensive. (Score:3, Informative)
That said, I have personally done the multi-seat thing, with the appropriate X patches (built into Ubuntu's x.org, had to patch Debian's xfree86) and the right hardware. I'm going to be deploying quad-seat machines to a small
Re:WAY too expensive. (Score:1)
Re:WAY too expensive. (Score:1)
This is a permissions problem, not a multi-seat issue.
Re:WAY too expensive. (Score:2)
Sounds like pulling access to a CLI would be required.
Thanks.
LoB
PCI-X (Score:4, Informative)
Why do you need an USB card? The server already comes with 2 USB ports, and an USB bus supports up to 127 devices.
Cat got your tongue? (Score:4, Interesting)
Re:Cat got your tongue? (Score:2)
Yup...the worst mistake I saw is the summary talking about the finished product supporting 28 users, when it states quite clearly in the software's webpage that the maximum number of users is 10.
Way to fact-check.
Matrox quad monitor PCI card (Score:2)
Here's a 4 port version on eBay:
Matrox G200 MMS G2+ QUAD PCI Video card/cable [ebay.com]
workstations - none, servers - quite a few (Score:5, Informative)
How many desktops is it practical for a high end server to directly replace?
None, just like a big truck doesn't replace any passanger cars.
You could, however try something like OpenVPS [openvps.org] to replace a couple 'o dosen servers with it...
Re:workstations - none, servers - quite a few (Score:2)
Repeat after me PCI-X != PCI Express (Score:3, Informative)
This machine has 0 AGP and 0 PCI Express slots, only "Graphics Integrated 1280 x 1024, 16M color on PCI local bus, 8 MB of SDRAM video memory".
Neat idea, but sorry.
Re:Repeat after me PCI-X != PCI Express (Score:3, Interesting)
PCI-X != PCI-Express (Score:4, Informative)
PCI-X is not PCI-Express, and the two technologies don't even have compatible pinouts or signals. PCI-X is the follow-on to traditional parallel PCI, with speeds of 100 and 133 Mhz and a 64-bit wide data path (compared to previous parallel PCI standards of 32/64-bits at 33/66 Mhz). PCI-Express is PCI re-done serially instead of in parallel, in an attepmt to be fashionable like the new Serial ATA standard. It's also potentially faster than PCI-X, and not at all compatible.
You'll notice just about every communications standard that doesn't go long haul alternates back and forth between parallel and serial methods every few years just to sound new and exciting and better.
Re:PCI-X != PCI-Express (Score:2)
I don't think this is true. In fact, what communications standards have alternated back and forth at all? I'm not a hardware guy, but I think the main expansion bus on desktop PCs (ISA/EISA/PCI) have always been parallel, right?
I think that most communications mediums, if they have alternated at all, have
Re:PCI-X != PCI-Express (Score:1)
ie more than one line but they aren't forced to run in precise lock-step
Re:PCI-X != PCI-Express (Score:2)
ie more than one line but they aren't forced to run in precise lock-step
That's what PCI-Express already does. Each card can use multiple "lanes", each lane is serial, and lanes are not synched. I don't know how many lanes the bus supports.
Re:PCI-X != PCI-Express (Score:2)
You'll notice just about every communications standard that doesn't go long haul alternates back and forth between parallel and serial methods every few years just to sound new and exciting and better.
That's awfully cynical. I don't think it's done to "sound new and exciting"; I think it's driven by the available technology. When advances are made that permit higher clock rates, we tend to see things shift toward serial interfaces; meanwhile, when such advances have not been made in a while, we tend t
600? (Score:1)
Re:600? (Score:2)
Re:600? (Score:2)
Re:600 users on one box = troll (Score:1)
2 gig limit per process?
You are smelling a lot like a troll
I/O advantages (Score:2)
VMWare (Score:2)
I know that the higher-end VMWare products support migrating between servers. I don't know how well they support multiple physical input and display devices, but I suspect that if a major customer requested it, it could happen
Marketing bait (Score:2)
SunRay + V480 (Score:3, Insightful)
The system was spec'd by Sun to handle those 100 sessions. The head engineer bought two and set them up to load-balance and provide redundancy.
This isn't anything new...move along.
Re:SunRay + V480 (Score:3, Informative)
For the "but the V480/V490 can't be ordered for $999 from Dell" trolls, there's also the v40z Opteron server that now sells with 4 dual-core CPUs.
However, for supporting a 100 desktops, something as robust as the V490 might be a good thing.
Re:SunRay + V480 (Score:1, Interesting)
Exactly. Why some people on here suggest using commodity parts to support a hundred users is beyond me. Perhaps they've never been in a business where if 100 people aren't working for 5 minutes they've just loss thousands of dollars of productivity.
It sounds like for $50k server + $30k clients = $80k you can get a robust system from sun. Throw in another $20k a year in maintaince and its only $100k for the fi
Re:SunRay + V480 (Score:3, Informative)
Sun is definitely a bargain, now, but they have to overcome the baggage of having the UltraSPARC II stuck at 480MHz while the UltraSPARC III was being delayed. That is the source of most of the "but my Pentium is five times faster for 1/20 the cost" trolls. Fortunately, that is _not_ a problem, anymore.
Re:SunRay + V480 (Score:1)
Re:SunRay + V480 (Score:2)
It looks like it's currently Solaris-SPARC and Linux-x86 as the supported configurations (yes, Slashdot, Sun _does_ sell Linux thin clients).
Re:SunRay + V480 (Score:2)
I hate to flog Windows, but.. (Score:5, Informative)
In the years since Buddy was first released, PCI video cards have learned to play nice with their neighbors, and USB has provided a way to connect oodles of keyboards and mice to the same machine. Thus, Buddy is reincarnated as BeTwin, a software-only product that associates specific keyboards, mice, and video cards with specific sessions on the machine. (I'm not sure how it deals with sound. Multiple soundcards would seem easy.)
They say it only supports 5 users, but that sounds like an arbitrary limit and I'm sure they'd tweak a 28-user version if you felt like it.
Related links... I'm going off-topic here, but playing stupid tricks with virtual hardware is fun.
Check out MaxiVista [maxivista.com], a "virtual video card" which Windows treats as a second monitor, allowing you to do multi-head tricks. The data for the second display goes out over the network (a la VNC [google.com]) to a client machine, which simply pipes it into the video buffer. Turn that scrap laptop into a second monitor! I stuffed a 10base-T card in my old lappy and it was perfectly usable for everything except fullscreen video. At 100 or gigabit, it'd be worth a try.
Xinerama [sourceforge.net] is Linux software that does the same thing, creating one large virtual X display, which then chops up the image and sends it to a number of smaller actual displays, some or all of which can obviously be located across the network.
As long as we're doing silly tricks with virtual hardware, you should be aware of Virtual Audio Cable [ntonyx.com], which enables digitally-perfect audio patching between applications' outputs and inputs, even if the apps themselves think they have exclusive control over the soundcard. (Also enables multiclient sound output under 9x, even if your card doesn't support it, because it does software mixing.)
If video is your thing, try Softcam [softcam.com], to feed your videoconferencing software any old source you feel like. Switch between actual cameras, use your desktop screenshot as a "camera input", add effects, etc. Their WaveMux tool is a nice complement to VAC, too.
Re:I hate to flog Windows, but.. (Score:2)
As I type I'm working on a PC with Thinsoft's "BeTwin [thinsoftinc.com]" software installed. Two video cards, two monitors, two keyboards, two mice -- two stations that you can log onto independently.
At work, I have Maxivista, though I haven't used it for a while since the power supply for a little network switch died. I must get that replaced.
Other interesting stuff to screw around with monitors includes; Margi's Display-to-go [margi.com] PCMCIA video ca
Ooh, USB2VGA, thanks for the tip. (Score:2)
Being USB 2.0, I'm surprised there's a bandwidth problem. I was running MaxiVista at 10Mbps, and it was tolerable. At USB1.1's 12Mbps, I'd expect similar. But at 480Mbps? It
Re:I hate to flog Windows, but.. (Score:2)
verb {T} -gg-
to beat someone very hard with a whip or a stick
(yes, I know you meant the British definition [freesearch.co.uk])
How Many Desktop PCs Can One Server Replace (Score:1, Informative)
Mainframe? (Score:5, Insightful)
As time went on and miniaturization progressed, people wanted their own department computer, so they could have more CPU time available.
Then they wanted their own desktops.
Then they wanted to network their desktop machines, so they could share data.
Then some applications started sharing CPU and other resources over the network.
But all these networked machines were a big configuration hassle for IT. They envisioned "thin clients" and similar solutions.
Now machines are so powerful that users can have their own virtual PCs running on a central server, so they can just have dumb terminals on their desks.
There's a lesson in here somewhere. As soon as the network comes back, I'll google for it and find out what that lesson is.
So lets do something a bit different. (Score:2)
Simplistically speaking you split the desktop computer into two. One is part of a "Big Server", the other is the "Thin Client".
If multicore CPUs and virtualization becomes common this isn't going to be that hard.
Of course if users randomly pull the plug on their nodes that does make things a bit problematic. So I suspect the current "thick desktop" stuff is going to be around for
Re:Mainframe? (Score:2)
No, they didn't. They didn't want to become system managers, they didn't want to have to spend their time defragging disks or dealing with viruses or any of those other problems.
A lot of managers found it easier to dump a bunch of cheap hardware on their staff and have them be their own system managers than to make a big up-front investment in a server and staff to deal with the server.
Only a few nerds preferred their own desktops. And some users that were saddled w
"most" traditional servers? (Score:2)
And there are plenty of those around.
We use lots of compute servers. After reliabaility, we care most about TCO per CPU/RAM set. A dual Opteron with 16GB is cheaper than two Opterons with 8GB each. But even if it was slightly more to buy, we'd take two in 1U over 2 in 2U. This scales forever, or until we hit som
Re:"most" traditional servers? (Score:2)
Indeed; the original question is how many PC's can a server replace but that's the wrong question. It should be "How many servers can a server replace?" Using VMWare, you can have what would otherwise be a rack full of little servers in one large machine. It costs less (when buying enterprise class hardware) and it's easier to manage. Dual core CPUs are a tremendous benefit when doing this.
I like the principle (Score:2)
I've thought about this a lot. Now that we have SMP and
Twenty-eight displays... (Score:3, Interesting)
PC-MOS / Multilink Advanced (Score:1)
Pretty cool stuff for the time. Also awesome for running multi-node BBS systems.