Linux equivalent to Cloud OS?

[angels355]

Hello,

I was wondering if anyone knows if there is a Linux version of Cloud OS? I've heard that Google and Yahoo are designing their own versions.

These type of OS's are as far as I know, a prepackaged OS system to operate super clusters of computers as one big super computer, like the previously known Beowulf super clusters.

In the past such super clusters have been put together with computers as weak as 25 mhz each. I have 300 mhz and upwards computers. Someone here has given me links to Beowulf how to websites and I'll have to look them up, but was wondering if there might be more that are good choices?

Thanks.

[azur]

You might be getting a bit mixed up with the latest buzzword...

It has other meanings, the the main one at the moment is effectively a marketing name for web 2.0. It's about having applications and data online, and available anywhere - "up in the clouds" so to speak. Gmail, remember the milk, blogs, facebook, etc. It's a pretty vague buzzword, but that's the idea.

So when sites talk about Google and Yahoo developing "cloud" operating systems, they're talking about desktop operating systems that use online based apps almost entirely. There was a very strong rumour that google would release one of these, and they have in the form of Android but it's for mobile platforms not the desktop. A linux based example of the concept is gOS, http://www.thinkgos.com/

If you're looking at attaching computers together, it doesnt work like 2x300mhz=600mhz - it's more like a multi-core computer. So to really use it, you need to use software designed to use those resources. An example on Linux is distcc for compiling.

[angels355]

I see what you're saying. I'll have to do more searching to see if there might be a more ready made OS for doing a beowulf as they (beowulf clusters) have been around for a while now. Appreciate your response.

[jwenting]

of course "web 2.0" is itself a marketing term :D
It was coined by O'Reilly as the name for a conference they organised about emerging technologies (itself a marketing term), and to their surprise it started appearing as a generic name for "interactive websites".

[angels355]

As a scientist I like marketing less, and actual technical advancement much more. I used to read about an Internet 2, which was supposed to be a super broad band internet, as they were saying that the present internet is not broad band enough and could soon be overwhelmed. And said that this new Internet 2 with super broad band could open up entirely new features to the public.

[jwenting]

people have been claiming that the internet will "run out of bandwidth" for almost as long as there's been an internet.
As it is the companies running the thing are constantly upgrading their hardware, so the internet of today is far different from the internet of even 5 years ago (let alone 30, which is when it all started with a few PDPs) and constantly morphing.

What will be needed soonish is the big move towards IPv6 from IPv4, but that's solely a software problem (albeit a massive one, as the majority of systems today don't even support IPv6, especially all those old Win98 and many WinXP and old Linux boxes used by FS9 addicts ;) ).

[angels355]

people have been claiming that the internet will "run out of bandwidth" for almost as long as there's been an internet.
As it is the companies running the thing are constantly upgrading their hardware, so the internet of today is far different from the internet of even 5 years ago (let alone 30, which is when it all started with a few PDPs) and constantly morphing.

What will be needed soonish is the big move towards IPv6 from IPv4, but that's solely a software problem (albeit a massive one, as the majority of systems today don't even support IPv6, especially all those old Win98 and many WinXP and old Linux boxes used by FS9 addicts ;) ).

How vulnerable is the internet to attack? I know a number of hombres mallos have threatened to either attack the internet itself or attack us through the internet. And will this IPv6 dramatically improve the security of the internet?


What will be needed soonish is the big move towards IPv6 from IPv4, but that's solely a software problem (albeit a massive one, as the majority of systems today don't even support IPv6, especially all those old Win98 and many WinXP and old Linux boxes used by FS9 addicts ;) ).

Wha-wha-wha..WHAT??? What are you impligrating?? This here computer runs 98se, very cutting edge I assure you! 500 mhz! Nothing less! I want to dual boot it with Ubuntu also but have to upgrade hdd's for more space. These days, if you wanna dual boot you gotta have more than a 4 gb hdd as your Master hdd! Lurnt that here just recently! :rolleyes: If I reckon, you were only running Ubuntu 6? Whew, ancient history! I know that Ubuntu 7.10 will run on my 500 mhz p3 Dell, I'm going to try out Ub 8.04.1 on it also as reluctantly it is nicer than 7.10. 7.10 however is a true classic, awesome! That is where it really did a great job of being able to read and write on windows partitions. Looking at 8.04.1 however, it is super nice. I switched hdd's from a 2.0 ghz celeron that was semi-modern, to a 600 mhz p3 Dell, and as NT5 & 5.1 do, if there's an important driver or set of drivers it won't run, and so I found that I either have to do a repair install or reinstall from scratch to get it to install the needed drivers for the downgrade. I erased the previous Ubuntu partitions and will soon reinstall one of the Ubuntus again from scratch. Surprisingly, I've found that both 7.10 and 8.04 will run on it, and 8.04.1 is very nice. 8.10 beta Ubuntu is out now, haven't tried it.

The nice thing about Linux is that it is constantly upgrading, I can hardly keep up with the upgrades. It seems like there's a new version of Ubuntu perhaps every six months I'm guestimating. Sabayon 3.5 the full DVD version is very nice also. It's amazing, it allows you to choose your interface for any given session, Gnome, KDE 3.5, xfce, and I think 2 or 3 others. With Linux I rest assured that if IPv6 comes out, that Linux will be ready for it, and w/o the $500 price tag for the OS.

One thing I do worry about however is the hardware. I use 300 to 500 mhz computers for the internet, and reserve my more powerful computers like 700 mhz to 933 mhz for 98se plus FS9 :cool:. Then my super duper computers from 1 ghz to 2.8 ghz for an intended win 2000 pro, plus FS9. It has been so useful to dual boot with Linux that I try to dual boot whenever possible. What's going to happen to this awesome over the top display of raw computing power when IPv6 comes out? Will a 500 mhz computer be able to handle it? And when it does upgrade to IPv6, will dialup work on it? Will IPv6 require universal broadband for everyone to even get on it?

[xxmikexx]

To my knowledge, at its lowest level today's TCP/IP is completely open and unprotected ...

There is nothing at that level, in principle, to prevent rogue external software from injecting stuff into your machine at that level. I believe it's only at the session level that security enters the picture. (Though maybe it's at the socket connect level. Been too long, I really don't remember the details. In fact, I'm no longer sure of even the terminology. :D )

[jwenting]

TCP, UDP, and IP are purely about transport and addressing, it does not, can not, and should not do filtering of any kind.

If you want security it has to be built on top of those protocols as an extra layer, or included in the underlying protocols or hardware.

That's by design, and it's a good design. There's a place for anything, and the IP layer is not the place for security.

[angels355]

Router hardware firewall is a good thing. And if you're lucky enough to be able to use Linux to access the internet, with far greater security protection, the learning curve to go from windows to linux for web surfing is about 2 seconds.

Thanks for the input Mike, jwenting's explanation of what these IP's, TCP's, and other hiccups are, is the first explanation of them that I've heard of. Not many people may know this, but my neighbor OWNS a Holiday Inn Express, so naturally I'm much more technically oriented as a result! :cool:

What about these super wide broad band experiments at universities? Has the present internet advanced to that level now? Or is there a super broad band coming in the near future?

W/ regards to electronic security, after the Cold War the electronics after that period were designed with lower standards and shielding with regards to protecting them from an electromagnetic pulse EMP. So it appears that we'd be wide open. (Program from the History Channel I believe.)

[jwenting]

The results of those experiments are trickling down into production hardware, but that takes time and with the decline in defense research budgets such research is becoming more scarce (a lot of it was funded at least in part by groups like DARPA, which are suffering badly as the military has trouble scraping enough money together to pay its people and maintain its equipment, let alone fight the war on terror).

But yes, EMP hardening of hardware is getting less as more and more of the core infrastructure is maintained and spec'd by corporations rather than governments.
Government agencies might still maintain a limited network to retain comms in case of massive failure, but it is nowhere near the percentage of the total that it was 20 years ago or even a decade ago.

[xxmikexx]

As I recall, the original DARPA purpose of the internet was to have a store-and-forward message system infrastructure that would survive nuclear attacks.

And yes, it was on PDPs at first. I was at DEC while one of our PDP-10s was part of the 25-node ARPAnet, though this was direct connect TTY protocol. (ARPA was the predecessor of DARPA.) It worked fine for 25 sites but the combinatorix would not have worked so well for 1,000 sites :D, another motivation behind the internet.

I had a summer intern write a program that bridged local PDP-11s into our Maynard, MA headquarters Arpanet PDP-10. I called it L-1011, L standing for link. So any site that wanted to buy 11-based communication front ends for its -10 would be able to have multiple users on any of the (dial-up) connections, depending on how many working modem boards were available at each end of any given connection. My failing memory says that this was the summer of 73, before we were able to integrate the comm front ends with the -10s to become true multiplexers as we know them today.

When I say "internet" I mean the early uses of TCP/IP. The whole point was to allow message packets to travel by any route available after which the message would be reassembled in correct packet order at the receiving end, as I recall.

[jwenting]

it was DARPA first, then ARPA, then DARPA again. Might have been more cycles depending on how politically correct it was to be openly associated with the DoD.

DARPANet grew out of a research net created by some systems engineers at a few universities (MIT and Stanford I believe), indeed with survivability as its core requirement.

[xxmikexx]

It was ARPA before DARPA, not the other way 'round ...

I was on ARPA projects from 1964 till I joined DEC in 71. As I recall, ARPA became DARPA around 75. Political correctness had nothing to do with the naming. It was simply that the military agencies wanted recognition that they were doing the funding, which had been true all along, ARPA being specifically military-oriented, though the research was conducted by civilians under contract. So they insisted on the D before the ARPA.

ARPA's military nature was not a secret. As a result we (Applied Data Research, ADR,a perennial ARPA contractor) got an inquiry in 67 or so from the Russian Academy of Sciences asking for a copy of a report on a certain research project into database structures. I don't think we sent it to them though the project leader wanted to do it.

As a member of that project for about six months, I struggled to invent what would later be called relational databases -- but I failed miserably. When I saw the solution many years later I smacked myself across the forehead because the correct ideas are really very simple.

We (ADR) did secret stuff too but that was for NSA -- No Such Agency -- and not ARPA. Because that was the case I had to get a Top Secret clearance even though I wasn't directly on any of the NSA projects. A lot of people around me were doing NSA contract work, and I heard stuff, and they asked me questions about certain technical matters, so I was effectively an adviser to certain of the technical teams. I can't say more because like everybody else I am still under a National Security oath that will never be lifted.

[jwenting]

it became ARPA again under Clinton because someone decided that it wouldn't do to have universities openly associated with the military.
By now that nonsense has I think been reverted again...

You're not the only one under NDAs, I don't think anyone in the industry hasn't had to sign one at some point over something (and usually the restrictions are worse in direct proportion to the unimportance of the work being done, only last month some colleagues had to sign one that mandated life in prison for disclosure about a project for youth services of a single city).

[xxmikexx]

Thanks for the update -- I didn't know about the name change back to ARPA.

It was not an NDA -- it was a sworn oath. To break a National Security oath is not an actionable civil offense, it's a federal crime. I don't recall signing anything, it will have been oral, memory fails, but I think it was at the huge FBI field office in Lower Manhattan, where I had been polygraphed months earlier when they began the background check for the T/S.

The background check was an interesting process. I don't recall whether I posted about it here or over at Train-Sim so I'll just do it again ...

xxxxxxxxxxxxxxxxxxxx

They started off by asking me about my political leanings. This was not funny because during the 30s my parents had been, quite literally, card carrying members of CPUSA. However, I was completely open about this, including about the fact that my mother's best friend had been one of Robert Oppenheimer's secretaries at Los Alamos, and about the fact that my parents had lived in Knickerbocker VIllage and had known Julius and Ethel Rosenberg.

I explained to the agents that I didn't share my parents' political views, and that my allegiance was to the USA, and that was that. Short of you yourself being a member of the Communist Party of the USA they really don't care what your politics are, or what other stuff may be lurking in your background. What they want to know is, Can You Be Blackmailed? If you tell them everything (repeat everything )then no, you can't be blackmailed, and that's the main qualification for getting a security clearance, that and US citizenship. The main purpose of the polygraph is to make sure that you're telling them everything (repeat everything).

During the course of that interview they asked me for the names of five friends who knew me well and had known me for several years. Later they asked each of those five people for the names of five people who might know me or know of me. Then they asked each of THOSE 25 people for five more names.

So before the elimination of duplicates they were prepared to do field interviews with 5+25+125 = 155 people who might know me or have heard of me. In about a month I started getting phone calls from friends and from friends of friends. They usually went like this: "I thought you would want to know that the FBI was here asking questions about you. Is everything okay? What's going on?" :D

xxxxxxxxxxxxxxxxxxxx

When I started at ADR in Princeton, NJ it was illegal to even mention the name of the agency in public. So if we were eating lunch in a restaurant and somebody wanted to say that he would be going to NSA headquarters the next day, he would say something like "I'm going to Fort Meade tomorrow. You know where to reach me." Or even simply "I'm going south tomorrow. I'll be driving."

[xxmikexx]

While I'm at it I might as well document a couple of post-scripts to what's above, so I'll have it in one place for when I sweep the material into my new blog at TheFreeSpeechForums.com ...

I kept running into the spooks over the years. At one point at DEC I had book/ship and manufacturing planning responsibility for the PDP-15 group, which included the business of its predecessor PDP-9 group. I got a call one day asking me to appear immediately at a meeting. At the meeting were two younger guys and one older guy, not DEC people but claiming to be DEC customers. They needed two PDP-9s immediately but wouldn't tell me what they were going to be used for. I explained that it would take months for us to build two -9s because they had been discontinued, but if they would just tell me a little bit more we would be able to divert a pair of -15s off the production line except for one except.

(I had already figured out who they had to be, had stepped out of the meeting briefly, and had made a phone call to our -15 production foreman. Not to his boss, who was a suit like me, but instead to Hank Buhlens, who ran the line itself from the shop floor, not from one of the offices upstairs and in the back. Hank told me the facts of life ...)

xxxxxxxxxxxxxxxxxxxx

So I said the following to the now obvious spooks ...

"You will need what are called 'fixed head' drives for those two machines. However, my production guy says that he can't ship the drives because because he's stocked out of the sealed bearings they need, and the vendor is quoting a six month lead time. Can you help us with that?"

"Of course" said one of the people. "We can get you anything you need. Have your guy call this number and explain what needs to happen."

We had the bearings by the next afternoon, and the machines left that night on a truck bound for a certain destination that I had already figured out for myself. Before the meeting ended I had drawn the senior guy aside and said "Look. I saw in the newspaper this morning that [whatever] happened [wherever], and I know now why you need the machines. But I need you to confirm it because you may need some special grounding arrangements and if so, our production people need to know that."

The senior guy confirmed my guess. I can't say more than that other than the fact that with the -9s in question being irrevocably down, the flow of information to Secretary of State Henry Kissenger regarding [whatever] was being seriously delayed, and it was of vital national concern that the replacement computers get on the air as soon as possible.

xxxxxxxxxxxxxxxxxxxx

About ten years later I invented a floppy disk redundant recording technique that I called "Datacure", for CP/M, an immediate predecessor of DOS for PCs. It was the equivalent of what would later be called RAID, except that the redundancy was built into the file structures on a single disk using a special driver that I wrote, rather than being spread across multiple drives. The result was that you could literally stick a pin into a Datacure floppy and get the data off, even though doing so might shorten the longevity of the read/write head.

Anyway, I took out some magazine advertising, and I took a booth at a trade show. The product went noplace in a sales sense but I did get a couple of interesting phone calls. The first one went like this ...

"Hi, my name is Bill Lastname. I would like to learn more about your product."
---------- "By any chance are you THE William Lastname, the inventor of the [whatever]?"
"Yes".
---------- "So why do you want to know? Are you representing somebody else?"
"Yes, I'm the deal screener for [names an investment group I had never heard of]. I saw your ad and decided this is something we should be aware of."

I explained to Lastname that there really was not a business here, that the technology was not proprietary though my use of it certainly was novel and possibly patentable, but that I would not be interested in their becoming investors since I was planning to pull the plug on the product and move on to other things.

A couple of days later I got another phone call of a similar nature. "We want you to send us an evaluation copy, we might need a bunch of them." He was quite evasive about who "we" was so I went out on a limb and said "Look, I know who you guys are. I'll tell you the same thing that I told Lastname a couple of days ago, and more. It works as follows. It does A, B, C and then D, and if there's a read error it then does Dprime, followed by Cprime, Bprime and Aprime. You don't have to buy anything, you don't have to be evasive, I'm giving you the information straight from the horse's mouth. If you need more information, have one of your people call back, I'll be happy to tell them whatever they want to know."

The guy laughed, confirmed that he couldn't tell me who he was for the very reasons I suspected :D, thanked me for the information, and I never heard from them again.

xxxxxxxxxxxxxxxxxxxxxxxxxx

But sometime in the mid-80s a book was published on the subject of NSA, one of many. Naturally I read them all, including this particular book, curious to see what it had to say about this, that and the other thing. Suddenly, deep into the book, I came to a pair of pages that laid out in black and white some information that should never have seen the light of day. Equally suddenly I realized two things.

First, somebody on the inside had spilled his guts to the book's author. I did then and do now consider that person to be a traitor.

Second, I suddenly understood why some things are classified for fifty and more years, and why other things are never declassified, including the information I possess relating to the two pages in question. As I now know, these kinds of restrictions are not capricious, not arbitrary, not the stupid whim of some ossified bureaucracy. They are vital to national security, and that's the reason for the oath that never expires.

[jwenting]

I agree some things should never see the light of day, at least not until all the people involved into the Xth generation are long dead.

I do however consider placing people under thread of life inprisonment without parole for "divulging information to a foreign national in times of war" when related to what is essentially a CRM application for a minor city subauthority to be a bit extreme :D

[xxmikexx]

jwenting,

Please elaborate. And what does CRM mean?

[jwenting]

CRM is customer relations management.
This does a bit more, but at the core that's what it is.

Basically it's a system that keeps track of which persons at which agencies in the city has contact with which people under the supervision of youth services.

Nothing secret about it, it allows for better coordination between doctors, police, youth services, etc. in the tracking of cases. The hope is this will make it easier to detect things like child abuse and youth gangs before they get out of control.

[xxmikexx]

Thanks, Jeroen

Now I see why secrecy might be a legitimate issue ...

This kind of thing is known in the intellegence community as "traffic analysis" -- who contacts whom with what frequency, and with what sudden changes in the patterns. So the software in question might be revealing of intelligence methods and therefore would want to be subject to NDA.

I'm guessing of course,
But it's an educated guess,