Beowulf is best for CPU intensive tasks which can be broken up easily, don't require a lot of intranode communication, can deal with relatively high latency on the intranode communication, and can deal with single node failures easily.
This is a relatively large domain or problems, but it doesn't work for everything. A lot of business applications require high reliability and availability. If you use beowulf, you have to implement these features for your application on your own.
The simulations that businesses are running on these things aren't really in the same league. For the most part, they aren't time critical and if a failure occurs that invalidates a test run, they can ususally be rolled back to some midpoint and started again without a significant loss of time.
Beowulf isn't just useful for CPU intensive tasks though. All those processors also provide significant amounts of memory bandwidth and all those machines provide potentially large amounts of disk storage and bandwidth, but again, you need memory or disk intensive tasks that can easily be split out to many loosely coupled nodes.
1. The article is in the Boston Globe, whose reputation is impeachable. Recall that a few of their star columnists and reporters were passing off fiction as fact.
2. The article relies on falun gong for the "facts" about these atacks originating from the Chineese government.
Consider the possibility that the Falun Gong are playing on anti-Chineese sentiment in the US in their efforts against the Chineese government.
Keep in mind that bandwidth isn't free. Without the ability to manage QoS you have a situation where the greedy can make things bad for everyone else.
This becomes a real issue when you are looking at high bitrate datastreams like streaming video. At todays prices, high quality video over the net is prohibitively expensive (8 mbs for TV quality high action video is not cheap).
That sort of backbone bandwidth can't be had for the price of a cable-modem connection, yet cable modem providers may want to offer such services. To do so at a price point that is affordable, they have to build local caching&reflection for the content, to keep traffic off their backbone connection.
The final peice is to allow high bandwidth streams between customers and distribution centers, where costs are well controlled, and disallow it between customers and the iternet at large.
If you don't like it, don't get a cable modem. Hell, if you don't like it, build your own local and global fibernetwork and give away access. If you can find a way to make it work, I will gladly contribute my time to making it happen.
I have no doubt that you could eventually build a reliable system from open source tools, but you should give serious consideration to a carrier class software like the stuff from software.com. I beleive their licensing starts in the 25,000 mailbox range and goes up, which should be a hint that they are ready to meet your needs.
Consider this 25,000 mailboxes * 10 MB/mailbox =250 GB. Is this something you are ready to trust to your linux filesystem without a second thought.
If you are going to roll your own you should dig around on netapp.com. They have a document which describes how earthlink built thier own e-mail system on top of a net-app filer.
You might view this as exclusionary, but it is a big step up from the old days when none of you peon-whiners would stand a chance at getting in on this or any other IPO.
You say "The government did not have to give away telephones and service or TV sets to have this staples of modern life become ubiquitous.
What you appearantly don't know is that the government has intervened in a variety of ways to make telephones as ubiquitous as possible, including forcing phone companies to operate unprofitable offices in exchange for the right to operate at all.
I know less about television, but I wouldn't be suprised if various steps were taken to trade transmission to remote areas for permission to operate, and possibly, for some protection from competition.
Interesting that you take this "Basically, all households, black and white, will soon be saturated by computers." away from the Cato article, because while it makes that assertion, it really does nothing to substantiate it.
The cato article states "Families that do not have computers now are going to have them in a few years. " but the evidence it gives does not support this conclusion.
Rather, it gives evidnce to support the idea that *many* Families that do not have computers now will have them in a few years. They try to show that the gap will close, but the truth is, they only show that there will be fewer people on the Have-not side of the gap than some doomsayers indicate.
These days people seem far too concerned about the way things look and not enough about how they work.
It used to be that people would just try to copy features of the Mac UI without fully understanding the thinking behind them, now it seems that even Apple isn't even doing that well.
I don't know that the traffic shaper device is appropriate. It seems that it only works on a subinterface basis, which suggests the need for a single sub-interface per subscriber.
This sounds like a waste of IP addresses to me. I would also worry about the ability of the kernel to support large numbers of subinterfaces in an efficient manner.
The previous poster suggests that you only throttle when bandwidth contention is an issue, suggesting that it will build goodwill.
I would suggest the opposite. You will certainly have a surplus of bandwidth when you roll out your service. If you open things up wide to everyone then you will probably have some very happy customers for a few months, and they will doubtless tell their friends. Soon you will have a growing customer base of people who are coming to expect more than they pay for. Then you have to start throttling down bandwidth. People are now getting less than they were getting before. Even if they are getting what they paid for, a lot of people are going to feel like they are getting shortchanged and they will start complaining vocally.
Maybe this isn't such a bad thing though. If you build a subscriber base quickly on word of mouth because you are giving away spare bandwith then you might be better off than if you build the subscriber base more slowly, or you have to advertise to build it quickly. It depends on how much the malcontents cost you once you have to start throttling connections, vs the costs of slower growth, or the costs of advertising. Unfortunately, the cost of the former is hard to predict.
As for dealing with the daily peaks of bandwidth utilization, again, I think people will tend to react better to consistant performance throughout the day or week, rather than wide fluctuations. On the other hand, if it is possible to allow maximal thruput on short (10-40k) spurts and throttle it down on longer downloas, then it becomes more difficult for people to quantify and less likely to engender ill will.
All this complaining that the OS & software used to set this record was tweaked!!
I don't see the problem. These people had access to source for AIX and they used it! Just the sort of thing that anyone with the skills can do to Open Source software.
Just because apache is incorporating doesn't mean they are on their path to becoming a big bad corporation. It doesn't mean an IPO is around the corner or any of that.
They are incorporating as a non-profit. This is not necessarily a good or a bad thing, but it does mean that people are doing what it takes to get to the next level.
Some of the outcomes I see.
1. It becomes easier for commercial companies to contribute. They may be able to write their cash and in kind contributions off on their taxes. At the very least, it should make their accounting a bit more straightforward.
2. It also makes it easier to take cash contributions. The ASF can take cash and do their own hiring, rather than working with donated time.
There are bad things that can happen as well, but they could happen anyway, so I am not going to bother listing them.
I am sorry, I could find little or no insight in the freshmeat article.
1. Microsoft admitting apache is better. -- I read this a different way. When a new software version is imminent Microsoft is not shy about ripping on the old one. You should have read the reviewers guides when Win95 came out, or talked to the SQL server guys at comdex last fall when MS-SQL 7.0 came out. They were absolutley merciless in their assement of the previous version.
With this context, I see this comment as more of a way of talking up windows 2000, the implication that its webserver will be as good or better than Apache. They may elevate the status of apache, but they say that they will have something as good or better than it.
2. Windows2000 will have a simple lightweight webserver -- Personal web server fits pretty much all of these criteria already, but I hardly think that anyone thinks that apache competes in this space anyway.
3. The new webserver will be available for addition to earlier microsoft operating systems. -- I guess we will see. I think that if microsoft thinks it is something special they will use it to sell the upgrade, rather than give it away. It is possible, I suppose, but I don't see anything that compelling.
4. Personal web server doesn't cut it and is hidden away a la IE2.0 -- does not compute. First. Personal web server isn't hidden away now. If it were much more out front it would be too intrusive. IE2.0 was on the desktop anyway. Beyond this, Personal web server is simple to use and relatively competant as is, I don't see how any current or future version would be targeted at apache.
5. The new MS web server may be crippled, but it will be used as an incentive to get a full version by upgrading to windows. Again, I don't see why this is relevant when we are talking about apache, not that it matters, because it already describes the status quo. Personal webserver is in some ways a first step to IIS and NT server.
This might make sense if the people who use apache as their first web server were the same sorts of people who would use Personal web server as their first web server, but I don't beleive they are.
6. Adding windows only features that integrate between this new web server and IE. Guess what, it has been happening since microsoft released their first web clients and servers. Never the less, I will speculate as to what would be a killer app for this. Full remote access to your machine through IE a la MS terminal server.
7. The new free web server / Windows 2000 combination will reduce the appeal of Apache. --- Huh, I don't see how, especially since this new free web server, by the articles description, is targeted at the low end.
Hey, I have another idea for a killer app. What if the new webserver integrated free microsoft provided web space so that you could work with things as if they were on your machine, but they were served to the world off of a big server farm with unlimited bandwidth. This would be hard for apache to compete with.
8. Attacking Apache with this kind of tactic is not legally painful as integrating the browser was. After all, Apache isn't made by a company. To the legal system and Microsoft, Apache came out of the ether and its secretary is a website. -- I dunno what this tactic could possibly be since I find this anatomy lesson to be remarkably short of meat.
9. how, exactly, is apache better than any microsoft web server in the sort of personal web server space you seem to be talking about. You haven't made your case.
10. The mindcraft benchmarks. What thinking person didn't realize that apache isn't everything that it has been built up to be? I think rational people were already choosing apache for different reasons.
Reading back over my argument, it doesn't seem to elgant, but then I was sparring with fantoms of the air anyway.
While your definition of ownership seems sensible enough, It doesn't hold up to the way things are done in the real world every day, especially in the world of intellectual property.
Slashdot Mirror
Beowulf is best for CPU intensive tasks which can be broken up easily, don't require a lot of intranode communication, can deal with relatively high latency on the intranode communication, and can deal with single node failures easily.
This is a relatively large domain or problems, but it doesn't work for everything. A lot of business applications require high reliability and availability. If you use beowulf, you have to implement these features for your application on your own.
The simulations that businesses are running on these things aren't really in the same league. For the most part, they aren't time critical and if a failure occurs that invalidates a test run, they can ususally be rolled back to some midpoint and started again without a significant loss of time.
Beowulf isn't just useful for CPU intensive tasks though. All those processors also provide significant amounts of memory bandwidth and all those machines provide potentially large amounts of disk storage and bandwidth, but again, you need memory or disk intensive tasks that can easily be split out to many loosely coupled nodes.
A few things to consider.
1. The article is in the Boston Globe, whose reputation is impeachable. Recall that a few of their star columnists and reporters were passing off fiction as fact.
2. The article relies on falun gong for the "facts" about these atacks originating from the Chineese government.
Consider the possibility that the Falun Gong are playing on anti-Chineese sentiment in the US in their efforts against the Chineese government.
This article is asinine, moderation is all well and good, but what can we do when editors post truly idiotic stories?
Such childishness.
Besides RC5 is boring!
Keep in mind that bandwidth isn't free. Without the ability to manage QoS you have a situation where the greedy can make things bad for everyone else.
This becomes a real issue when you are looking at high bitrate datastreams like streaming video. At todays prices, high quality video over the net is prohibitively expensive (8 mbs for TV quality high action video is not cheap).
That sort of backbone bandwidth can't be had for the price of a cable-modem connection, yet cable modem providers may want to offer such services. To do so at a price point that is affordable, they have to build local caching&reflection for the content, to keep traffic off their backbone connection.
The final peice is to allow high bandwidth streams between customers and distribution centers, where costs are well controlled, and disallow it between customers and the iternet at large.
If you don't like it, don't get a cable modem. Hell, if you don't like it, build your own local and global fibernetwork and give away access. If you can find a way to make it work, I will gladly contribute my time to making it happen.
Of course outlook doesn't scale, it is an e-mail client, not an e-mail server.
Perhaps you meant exchange server. If you did, I would have to agree with you. Exchange server would be wrong for this sort of arangement.
I have no doubt that you could eventually build a reliable system from open source tools, but you should give serious consideration to a carrier class software like the stuff from software.com. I beleive their licensing starts in the 25,000 mailbox range and goes up, which should be a hint that they are ready to meet your needs.
Consider this 25,000 mailboxes * 10 MB/mailbox =250 GB. Is this something you are ready to trust to your linux filesystem without a second thought.
If you are going to roll your own you should dig around on netapp.com. They have a document which describes how earthlink built thier own e-mail system on top of a net-app filer.
You might view this as exclusionary, but it is a big step up from the old days when none of you peon-whiners would stand a chance at getting in on this or any other IPO.
If you can't prioritize your spending to afford $17/mo on your kids, then you shouldn't be having kids in the first place.
So, what other criteria should we employ when deciding whether people should have kids, eh, SpinyNorman?
You say "The government did not have to give away telephones and service or TV sets to have this staples of modern life become ubiquitous.
What you appearantly don't know is that the government has intervened in a variety of ways to make telephones as ubiquitous as possible, including forcing phone companies to operate unprofitable offices in exchange for the right to operate at all.
I know less about television, but I wouldn't be suprised if various steps were taken to trade transmission to remote areas for permission to operate, and possibly, for some protection from competition.
Interesting that you take this "Basically, all households, black and white, will soon be saturated by computers." away from the Cato article, because while it makes that assertion, it really does nothing to substantiate it.
The cato article states "Families that do not have computers now are going to have them in a few years. " but the evidence it gives does not support this conclusion.
Rather, it gives evidnce to support the idea that *many* Families that do not have computers now will have them in a few years. They try to show that the gap will close, but the truth is, they only show that there will be fewer people on the Have-not side of the gap than some doomsayers indicate.
I would care a whole lot more if someone would sell a nice quite case. I am really getting sick of all the whiring noises in my offices.
Moderator: Funny? About as funny as having your kid's slaughtered in front of you.
Amiga is still a gateway owned company, right?
These days people seem far too concerned about the way things look and not enough about how they work.
It used to be that people would just try to copy features of the Mac UI without fully understanding the thinking behind them, now it seems that even Apple isn't even doing that well.
Given the cost of a cisco, it makes sense to consider doing the QOS stuff on another box.
I don't know that the traffic shaper device is appropriate. It seems that it only works on a subinterface basis, which suggests the need for a single sub-interface per subscriber.
This sounds like a waste of IP addresses to me. I would also worry about the ability of the kernel to support large numbers of subinterfaces in an efficient manner.
The previous poster suggests that you only throttle when bandwidth contention is an issue, suggesting that it will build goodwill.
I would suggest the opposite. You will certainly have a surplus of bandwidth when you roll out your service. If you open things up wide to everyone then you will probably have some very happy customers for a few months, and they will doubtless tell their friends. Soon you will have a growing customer base of people who are coming to expect more than they pay for. Then you have to start throttling down bandwidth. People are now getting less than they were getting before. Even if they are getting what they paid for, a lot of people are going to feel like they are getting shortchanged and they will start complaining vocally.
Maybe this isn't such a bad thing though. If you build a subscriber base quickly on word of mouth because you are giving away spare bandwith then you might be better off than if you build the subscriber base more slowly, or you have to advertise to build it quickly. It depends on how much the malcontents cost you once you have to start throttling connections, vs the costs of slower growth, or the costs of advertising. Unfortunately, the cost of the former is hard to predict.
As for dealing with the daily peaks of bandwidth utilization, again, I think people will tend to react better to consistant performance throughout the day or week, rather than wide fluctuations. On the other hand, if it is possible to allow maximal thruput on short (10-40k) spurts and throttle it down on longer downloas, then it becomes more difficult for people to quantify and less likely to engender ill will.
All this complaining that the OS & software used to set this record was tweaked!!
I don't see the problem. These people had access to source for AIX and they used it! Just the sort of thing that anyone with the skills can do to Open Source software.
And why did cringley decline?
The level of paranoia here is amazing.
Just because apache is incorporating doesn't mean they are on their path to becoming a big bad corporation. It doesn't mean an IPO is around the corner or any of that.
They are incorporating as a non-profit. This is not necessarily a good or a bad thing, but it does mean that people are doing what it takes to get to the next level.
Some of the outcomes I see.
1. It becomes easier for commercial companies to contribute. They may be able to write their cash and in kind contributions off on their taxes. At the very least, it should make their accounting a bit more straightforward.
2. It also makes it easier to take cash contributions. The ASF can take cash and do their own hiring, rather than working with donated time.
There are bad things that can happen as well, but they could happen anyway, so I am not going to bother listing them.
I am sorry, I could find little or no insight in the freshmeat article.
1. Microsoft admitting apache is better. -- I read this a different way. When a new software version is imminent Microsoft is not shy about ripping on the old one. You should have read the reviewers guides when Win95 came out, or talked to the SQL server guys at comdex last fall when MS-SQL 7.0 came out. They were absolutley merciless in their assement of the previous version.
With this context, I see this comment as more of a way of talking up windows 2000, the implication that its webserver will be as good or better than Apache. They may elevate the status of apache, but they say that they will have something as good or better than it.
2. Windows2000 will have a simple lightweight webserver -- Personal web server fits pretty much all of these criteria already, but I hardly think that anyone thinks that apache competes in this space anyway.
3. The new webserver will be available for addition to earlier microsoft operating systems. -- I guess we will see. I think that if microsoft thinks it is something special they will use it to sell the upgrade, rather than give it away. It is possible, I suppose, but I don't see anything that compelling.
4. Personal web server doesn't cut it and is hidden away a la IE2.0 -- does not compute. First. Personal web server isn't hidden away now. If it were much more out front it would be too intrusive. IE2.0 was on the desktop anyway. Beyond this, Personal web server is simple to use and relatively competant as is, I don't see how any current or future version would be targeted at apache.
5. The new MS web server may be crippled, but it will be used as an incentive to get a full version by upgrading to windows. Again, I don't see why this is relevant when we are talking about apache, not that it matters, because it already describes the status quo. Personal webserver is in some ways a first step to IIS and NT server.
This might make sense if the people who use apache as their first web server were the same sorts of people who would use Personal web server as their first web server, but I don't beleive they are.
6. Adding windows only features that integrate between this new web server and IE. Guess what, it has been happening since microsoft released their first web clients and servers. Never the less, I will speculate as to what would be a killer app for this. Full remote access to your machine through IE a la MS terminal server.
7. The new free web server / Windows 2000 combination will reduce the appeal of Apache. --- Huh, I don't see how, especially since this new free web server, by the articles description, is targeted at the low end.
Hey, I have another idea for a killer app. What if the new webserver integrated free microsoft provided web space so that you could work with things as if they were on your machine, but they were served to the world off of a big server farm with unlimited bandwidth. This would be hard for apache to compete with.
8. Attacking Apache with this kind of tactic is not legally painful as integrating the browser was. After all, Apache isn't made by a company. To the legal system and Microsoft, Apache came out of the ether and its secretary is a website. -- I dunno what this tactic could possibly be since I find this anatomy lesson to be remarkably short of meat.
9. how, exactly, is apache better than any microsoft web server in the sort of personal web server space you seem to be talking about. You haven't made your case.
10. The mindcraft benchmarks. What thinking person didn't realize that apache isn't everything that it has been built up to be? I think rational people were already choosing apache for different reasons.
Reading back over my argument, it doesn't seem to elgant, but then I was sparring with fantoms of the air anyway.
While your definition of ownership seems sensible enough, It doesn't hold up to the way things are done in the real world every day, especially in the world of intellectual property.
Give nothing? How about web hosting for people to chintzy to pay for their own bandwidth?
Don't attribute to malce that which can easily be explained by stupidity.