'Supercomputer on a CD' software is supposed to make it so easy to put a Beowulf cluster together
I've wanted to build a cluster for a long time. I was given that chance at school & work. School(U of C) allowed me to do it as a term project that is still ongoing. Work allowed me to use a pile of spare machines that were waiting for new users.
So, early one Saturday morning, I sit down with the O'Reilley book "Building Linux Clusters" and the CD that came with it. I followed the instructions in the book, and was frustrated beyond belief. The CD contained all the Beowulf software that was required. The downside was that the software had all been thrown over top of Red Hat 6.2. Being the second time I had ever installed Red Hat, I wasn't sure what the magic sequence was to get everything to work.
I first had problems with unsupported video cards... I tried 3 different cards. Each time, I needed to re-install Red Hat. Why wouldn't it let me install all the drivers for all the cards?
DHCP? Why? You only need to set the IPs once. Don't force me to do this... oh wait, I don't know Red Hat's weird config script structure... sigh.
The book mentioned nothing about re-compiling the kernel. But, in order to add support for the network cards, this is what I needed to do. Oh wait, where are the kernel sources?
This is when I got sick of this "wonderful" Beowulf CD.
I went to Slackware 7.1.
Installed it on the master no problem. Enabled frame buffer support for the video card so that it would work on ANY video card. Enabled native support for the network cards I was using.
Next step, I went to THE beowulf site, did a search for PVM and PVMPOV. I downloaded all the source code I needed.
Now, without the help of the book, I was at a bit of a loss. Luckily, there was this site that explained EVERY STEP in about a page and a half.
The how-to was written by Christopher Johnson and I must say, he did an excellent job. I found only one thing that was lacking, you may also need to set PVM_DPATH=/home/pvm3/lib/pvmd in your profile.
Now the purpose behind all of this was to get PVMPOV running, well, with a little searching, I was able to find everything I needed here.
Conclusions:
Use a Linux distro you are used to.
Get a book if you want to know the theory.
Always remember that some PVM Books are free.
I hope this will help someone out there.
Beware TPB
Vote for wether or not it should be cloned...
on
TigerCloning
·
· Score: 1
Currently in the office I administer, we use Office 2K. Not because we like it, or because it's stable, but because it can do the job and it's (relatively) easy to use. If there was a cheaper alternative that allowed us to read/write micosoft DOC and XLS, we would probably use that instead. At $800 (Canadian) per license of Office, it's far too expensive for what you get.
I think now is the time for big groups like these to start marketing their wares. I mean, a little while ago, the world "suddenly" found out that Micosoft is evil... now people are starting to look for alternatives. Linux has begun to reach a certain point of user-friendlyness that makes the OS itself usable in buisness. All that now remains is a solid set of apps for "productivity".
In my opinion, this project has a much better chance of success than any of the others thus far. When the new products become available, I for one will show them to the people I support in hope of having them switch over. The cost difference alone might be enought to convince them.
The "average" Americain has an attention span of about 30 minutes (based on my own experience). This is why they can't handle most anime. It's the same reason many of them don't listen to classical music. If they have to remember what happened in the last episode, they're going to loose interest. It is a sad statement, but it is true.
The stereotypical male Americain works nine to five, five days a week, and spends their weekend sitting infront of football on the TV with a beer in one hand and the remote in the other. They're not looking for plot development, they're watching to see people get injured.
Knowing this, the Hollywood cookie-cutter produces a thousand shows & movies based on the same idea: people getting blown up or slaughtered will appeal to the lowest common denominator.
Anime, on the other hand relies on long-term story lines, plot development, character development, and the viewer being able to remember past events and subtleties.
So, odds are if you like an epic tale, you'll always be sadly disappointed by Americain television.
Now what happens when Intel decides to release the "Pentium 5"? I mean, I still call the Pentium a "P5". What they should do is just use the damn "in house" code names. It would make just as much sense.
I should go out and buy a P6, then when the Willamete (or whatever) is released, I can say: "Hey, for $3000 I'll sell you my Pentium Pro, it's better than a P4 because it's a PRO!" Then get $12,000 from some EBay junkie.
I know very little of system administration, but the way I see it, the only way to transmit spam, is to have an ISP that is either incompetent, or willfully allowing it. If I'm not mistaken, sendmail will (by default) not allow external relay access. These laws should make ISPs responsible, then you would see a drastic decline in levels of spam.
Then, the problem becomes inforcing the laws in other nations. The Internet needs some sort of internal justice system. No single country should be able to write laws like this. These laws have to come from the users, wherever they are *physically* located.
This "Internet Justice System" would only be feasible if it had the power to deny service to certain ISPs and countries based on their willingness/unwillingness to comply with the rules brought forth by the Users.
Anyway, that's my rant for today, let me know what your thoughts are on the idea.
There are nine basic things that make a good interface. (begin Plug) This was part of the material covered in a class I took at the university of Calgary under Saul Greenberg (CPSC481). It's a fairly good course if you're in the area. (End of Plug)
Be consistent - from screen to screen and between your UI and the OS
Provide Help - online help and navigation ques to help the user along
Simple and natural dialogue - this involves a good grouping of items in a dialog and a layout that makes sense to the USER, not you.
Speak the users' language - The software will do no good if it uses terms that the user doesn't know.
Minimize user memory load - Relevant information should be carried forward from one screen to the next.
Provide feedback - If the user pushes a button and the computer is doing some complex operation, let the user know (roughly) how long it will take, and what it's doing.
Provide clearly marked exits - Have you ever been unable to get out of a piece of software?
Provide shortcuts - Although not crutial, for repetitive tasks these are nice.
Deal with errors in positive and helpful manner - do not insult the user, tell them what's wrong in terms that they can understand, and tell them how to fix the problem.
I'm now finishing my third year of my Computer Science degree. In our second year, we studied PDP-11 (in one course) for a full term. I can program machine code (zeros and ones) on the PDP-11. The rest of my class can also do this. Man, the PDP-11 was the most amazing chip (or so it seamed).
Flat memory model Orthogonal instruction set
Those were the days. I went in knowing how to code x86 assembly, by the time I was done, I couldn't read it anymore.
Are there any modern chips out that have an orthogonal instruction set and a flat memory model that can do 64bit integer operations? I would dearly love to play with such a beast.
Slashdot Mirror
'Supercomputer on a CD' software is supposed to make it so easy to put a Beowulf cluster together
I've wanted to build a cluster for a long time. I was given that chance at school & work. School(U of C) allowed me to do it as a term project that is still ongoing. Work allowed me to use a pile of spare machines that were waiting for new users.
So, early one Saturday morning, I sit down with the O'Reilley book "Building Linux Clusters" and the CD that came with it. I followed the instructions in the book, and was frustrated beyond belief. The CD contained all the Beowulf software that was required. The downside was that the software had all been thrown over top of Red Hat 6.2. Being the second time I had ever installed Red Hat, I wasn't sure what the magic sequence was to get everything to work.
I first had problems with unsupported video cards... I tried 3 different cards. Each time, I needed to re-install Red Hat. Why wouldn't it let me install all the drivers for all the cards?
DHCP? Why? You only need to set the IPs once. Don't force me to do this... oh wait, I don't know Red Hat's weird config script structure... sigh.
The book mentioned nothing about re-compiling the kernel. But, in order to add support for the network cards, this is what I needed to do. Oh wait, where are the kernel sources?
This is when I got sick of this "wonderful" Beowulf CD.
I went to Slackware 7.1.
Installed it on the master no problem. Enabled frame buffer support for the video card so that it would work on ANY video card. Enabled native support for the network cards I was using.
Next step, I went to THE beowulf site, did a search for PVM and PVMPOV. I downloaded all the source code I needed.
Now, without the help of the book, I was at a bit of a loss. Luckily, there was this site that explained EVERY STEP in about a page and a half. The how-to was written by Christopher Johnson and I must say, he did an excellent job. I found only one thing that was lacking, you may also need to set PVM_DPATH=/home/pvm3/lib/pvmd in your profile.
Now the purpose behind all of this was to get PVMPOV running, well, with a little searching, I was able to find everything I needed here.
Conclusions:
Use a Linux distro you are used to.
Get a book if you want to know the theory.
Always remember that some PVM Books are free.
I hope this will help someone out there.
Beware TPB
Vote!
Currently in the office I administer, we use Office 2K. Not because we like it, or because it's stable, but because it can do the job and it's (relatively) easy to use. If there was a cheaper alternative that allowed us to read/write micosoft DOC and XLS, we would probably use that instead. At $800 (Canadian) per license of Office, it's far too expensive for what you get.
I think now is the time for big groups like these to start marketing their wares. I mean, a little while ago, the world "suddenly" found out that Micosoft is evil... now people are starting to look for alternatives. Linux has begun to reach a certain point of user-friendlyness that makes the OS itself usable in buisness. All that now remains is a solid set of apps for "productivity".
In my opinion, this project has a much better chance of success than any of the others thus far. When the new products become available, I for one will show them to the people I support in hope of having them switch over. The cost difference alone might be enought to convince them.
ttyl
tpb
The "average" Americain has an attention span of about 30 minutes (based on my own experience). This is why they can't handle most anime. It's the same reason many of them don't listen to classical music. If they have to remember what happened in the last episode, they're going to loose interest. It is a sad statement, but it is true.
The stereotypical male Americain works nine to five, five days a week, and spends their weekend sitting infront of football on the TV with a beer in one hand and the remote in the other. They're not looking for plot development, they're watching to see people get injured.
Knowing this, the Hollywood cookie-cutter produces a thousand shows & movies based on the same idea: people getting blown up or slaughtered will appeal to the lowest common denominator.
Anime, on the other hand relies on long-term story lines, plot development, character development, and the viewer being able to remember past events and subtleties.
So, odds are if you like an epic tale, you'll always be sadly disappointed by Americain television.
ttyl
tpb
airbags would inflate to cushion the 50-mph landing.
So... I take it they still havn't decided to stick with Metric... I can see the headlines now...
50-mph? I thought you meant 50Km/h! Oh crap!
ttyl
tpb
80486 -> 486
P5 -> Pentium
P6 -> Pentium Pro
Now what happens when Intel decides to release the "Pentium 5"? I mean, I still call the Pentium a "P5". What they should do is just use the damn "in house" code names. It would make just as much sense.
I should go out and buy a P6, then when the Willamete (or whatever) is released, I can say: "Hey, for $3000 I'll sell you my Pentium Pro, it's better than a P4 because it's a PRO!" Then get $12,000 from some EBay junkie.
just my $.02.
Beware TPB
We've had the technology for close to ten years now. It's not going to happen because the people in power fear the technology.
On a side note, this would probably bring us one step closer to Direct Democracy instead of this whole "elected representation" thing.
Beware TPB
Do you use MPI? Do you find it restrictive in any way? Have you ever thought about using your own message passing routines?
ttyl
tpb
I know very little of system administration, but the way I see it, the only way to transmit spam, is to have an ISP that is either incompetent, or willfully allowing it. If I'm not mistaken, sendmail will (by default) not allow external relay access. These laws should make ISPs responsible, then you would see a drastic decline in levels of spam.
Then, the problem becomes inforcing the laws in other nations. The Internet needs some sort of internal justice system. No single country should be able to write laws like this. These laws have to come from the users, wherever they are *physically* located.
This "Internet Justice System" would only be feasible if it had the power to deny service to certain ISPs and countries based on their willingness/unwillingness to comply with the rules brought forth by the Users.
Anyway, that's my rant for today, let me know what your thoughts are on the idea.
ttyl
TPB
What happens if a song says "Next Track"?
There are nine basic things that make a good interface. (begin Plug) This was part of the material covered in a class I took at the university of Calgary under Saul Greenberg (CPSC481). It's a fairly good course if you're in the area. (End of Plug)
Be consistent - from screen to screen and between your UI and the OS
Provide Help - online help and navigation ques to help the user along
Simple and natural dialogue - this involves a good grouping of items in a dialog and a layout that makes sense to the USER, not you.
Speak the users' language - The software will do no good if it uses terms that the user doesn't know.
Minimize user memory load - Relevant information should be carried forward from one screen to the next.
Provide feedback - If the user pushes a button and the computer is doing some complex operation, let the user know (roughly) how long it will take, and what it's doing.
Provide clearly marked exits - Have you ever been unable to get out of a piece of software?
Provide shortcuts - Although not crutial, for repetitive tasks these are nice.
Deal with errors in positive and helpful manner - do not insult the user, tell them what's wrong in terms that they can understand, and tell them how to fix the problem.
that about sums it up. have fun
I'm now finishing my third year of my Computer Science degree. In our second year, we studied PDP-11 (in one course) for a full term. I can program machine code (zeros and ones) on the PDP-11. The rest of my class can also do this. Man, the PDP-11 was the most amazing chip (or so it seamed).
Flat memory model
Orthogonal instruction set
Those were the days. I went in knowing how to code x86 assembly, by the time I was done, I couldn't read it anymore.
Are there any modern chips out that have an orthogonal instruction set and a flat memory model that can do 64bit integer operations? I would dearly love to play with such a beast.