Microsoft Finally up for Distributed Computing? 307
ReeprFlame writes "eWeek has reported overhearing Microsoft's plans to finally get into the distributed computing market. Considering that the Windows platform has never had the ability to parallel compute in the past, it leaves great potential to the company's operating system development. From current *nix systems we have today, such a grid proves very useful, especially in the serving arena. However, we are unsure of Microsoft's target for the software. Would it be an addition to home users computers as well as the server versions of Windows? As of now it is unclear, but Microsoft probably will bring this situation to life in the near future since it does hold alot of power for them over other platforms."
Sun GridEngine (Score:2, Informative)
- Windows XP and 2000 (December 2004 availability)
http://www.sun.com/software/gridware/ [sun.com]
Gridengine's source can be downloaded from:
http://gridengine.sunsource.net/ [sunsource.net]
hardware is the cost (Score:3, Informative)
On linux you can remove interrupts from the kernel if your app only needs polling. Stuff like that will never be possible with a closed source solution.
Lots of ppl stop using solaris cause of this.
confusing parallel and distributed computing (Score:5, Informative)
distributed computing happens at the application layer. Thus if you can run something like an MPI library on windows you have the basis for efficient distributed computing. All you need is a scheduler and launcher to be able to launch distributed launch an application across the net. But virtually all of these are daemons not strictly part of the OS. So that level of system independent abstraction exists already so this should not be too difficult.
No ETs yet... (Score:4, Informative)
Of note: I've got some Win2K web servers running in a native WLBS load balanced rig, and those machines have been doing swell for four years now. They talk to a cluster of SQL servers, but that clustering really doesn't count... it's more like hot fail-over. The native load balancing of the web servers, though, has been pretty tight and has scaled very easily, at least within my mid-market universe.
I know, I'm just asking for it with this post. Just wanted folks to know that it's possible to push a couple $million of holiday e-commerce through some pretty cheap white boxes running MS's stuff. And yes, my cheap admin help is glad there's a GUI for some of the chores they don't do every day. All right, flame me now. But you have to do it from a command prompt.
Re:Who wrote the summary? (Score:5, Informative)
A Peek Under Microsoft's Secret 'Bigtop'
By Mary Jo Foley, Microsoft Watch
December 29, 2004
Microsoft officials have said little about the company's intentions in the grid-computing space. But that doesn't mean Microsoft is ignoring the evolving arena of grid/distributed computing.
Microsoft is working on a skunk-works project that is code-named Bigtop, which is designed to allow developers to create a set of loosely coupled, distributed operating-systems components in a relatively rapid way, according to sources close to the company, who requested anonymity.
Rather than attempting to tightly couple a few high-performance systems together, Microsoft is looking at the consequences of loosely coupling a larger number of moderately powerful computers to achieve a similar result.
Bigtop's first commercial manifestation will likely be as some kind of large-scale project, most likely a distributed grid-computing operating system, the sources added.
Bigtop is one of Microsoft's incubation projects. It falls under the domain of Craig Mundie, the Microsoft senior vice president and chief technical officer in charge of advanced strategies and policy, sources said.
Bigtop consists of three components, all written in C#, according to developers who said they were briefed by Microsoft. These are:
Highwire: Highwire is a technology designed to automate the development of highly parallel applications that distribute work over distributed resources, the aforementioned sources said. Highwire is a programming language/model that will aim to make the testing and compiling of such parallel programs much simpler and more reliable.
Bigparts: Bigparts is code designed to turn inexpensive PC devices into special-purpose servers, according to the sources. Bigparts will enable real-time, device-specific software to be moved off a PC, and instead be managed centrally via some Web services-like model.
Bigwin: According to sources close to Microsoft, Bigwin sounds like the ultimate manifestation of Microsoft's "software as a service" mantra. In a Bigwin world, applications are just collections of OS services that adhere to certain "behavioral contracts." These OS services can be provided directly by the core OS or even obtained from libraries outside of the core OS.
Sources said Microsoft will likely make some sort of preview version of the Bigtop code available to the company's software-development partners by 2006. If and when the final version debuts, it won't be much before the end of the decade, sources added.
It's not clear whether the Bigtop components will run on top of Windows when they are completed. But sources say that is what they are expecting at this point. End of Article
I like their use of a circus term as a name for this project. It gives the impression of a bunch of clowns running around into each other and falling down. Kind of like MS systems on the web now.
Re:Windows clusters don't make sense (Score:5, Informative)
I still feel that Linux would be a good bit cheaper, but we're talking big bucks both ways. And it's also worth mentioning that Microsoft's licensing model for "corner cases" like this is extremely flexible: they may give the source away at a significant discount just for the publicity. They've done it plenty of times before. Some of those 2500 servers at work run a custom-built NT kernel and we sure aren't a huge international company.
Re:Windows clusters don't make sense (Score:0, Informative)
Do everyone a favor and do some RESEARCH before you go spouting-off about something that you obviously know NOTHING about.
For some enlightenment, go to www.windowsclusters.org [windowsclusters.org]
--ScottKin
Not entirely accurate (Score:1, Informative)
This isn't entirely accurate. Server 2003 supports clustering, and if I remember correctly it even has a toolkit that allows you to add XP Pro systems to a cluster started by the Server 2003 machine. A friend of mine created a small cluster as part of one of his courses last year.
Re:Not entirely accurate (Score:3, Informative)
MS like to call it a cluster because it makes them sound 'good', but really it's crap.
Re:Windows clusters don't make sense (Score:3, Informative)
Even if I had a SSH/telnet-driven command prompt, I don't think I could kill a process on a remote machine, for example; I can do it only via the GUI. Is it just because I have a lot to learn, or is it a feature I don't have?
rkill, but I think it's an installable service that only comes with Resource Kit.
Repeat After Me (Score:2, Informative)
Windows does not have clustering!
Although they may have the capability of real clustering some day they do not have that capability today no matter how much your resident MCSE talks about his great exchange clusters etc. Windows can load balance and it can provide failover and it can run some distributed processing software but it cannot natively cluster.
Linux on the other hand has the tools available to run a true cluster, failover, load balancing and a real cluster aware file system meaning all nodes can share, distribute and balance processes. A clustered file system means that all processes even running on cluster nodes can access the same exact data.
For the linux crowd I would not exactly worry about this as MS is light years behind when it comes to this capability. We run some CFD solvers at work which where initially put on some 2K boxes and have been since migrated to linux to eliminate system crashes and improve the solution speed by nearly 50%.
We do have some oracle rac installations but those I don't consider real clusters either as the database file system clustering is not general purpose like Luster or GFS.
Re:since when are programs ran when they're not us (Score:3, Informative)
Yes, they're in the resourcekit for windows 2000/2003/XP (and a lot of other command line tools)
But you don't have to do this, you can for example remotely login using terminal services for admin usage, even if the server doesn't have monitor,mouse and keyboard attached. But if you want to config windows using a commandline, you can.
rescanning hardware changes can be triggered by a reboot.
Re:Windows clusters don't make sense (Score:3, Informative)
The Windows GUI can be turned off, along with many of the other services that you won't need in a cluster. It is not even that hard. The MS knowledge base is a mess, but the information is there. There are many performance tweaks for the NT kernel that don't require a recompile. It should be noted that most Linux clusters use unmodified, or lightly modified kernels. Most admins feel that the slight performance gain (if any) is not worth the maintanance. While a Linux license could cost $0 most clusters pay more. The cluster owners want the maintanance that comes with a commercial Linux distro. The Windows licenses are actually not that expensive, and MS is also more then willing to negotiate a better price. Linux would probably cost less, but not that much less.
Finally there are some nice tools for Windows that allow you to manage a large cluster. Tools that are more sophisticated then ssh or distributed ssh. Tools similar to the ones that RedHat and Suse are busilly writing because the equivalent Linux tools are pretty basic.
There are many reasons why Linux is perfect for clusters. But none of the reasons you list are valid.
Microsoft Cluster Server (MSCS) (Score:2, Informative)
Clustering Solutions for Windows NT:
http://www.windowsitpro.com/Windows/Article/Art
http://www.amazon.com/exec/obidos/tg/detail/-/0
http://research.microsoft.com/users/gray/Wolfpa
I can't be the only one that had this book.
BRING BACK THE TROLLS!!! (Score:3, Informative)
No, I don't think that's the reason. Metamod takes care of clueless mods. I don't know about others, but the main reason why I read much less Slashdot than I used to is that trolls have been effectively defanged.
Slashdot was a really funny site to read when one could find humorous, although often off-topic, gems of internet wisdom inside. Most of the posts that get modded "funny" today would be better classified as "trite" or "corny". Today, the people who made Slashdot what it was have been banned. No more Natalie Portman, no more goatse ascii, no more anything that's even remotely diferent from the mainstream media.
Well, for mainstream media we have Geraldo and Larry King, thanks, we don't need Slashdot for that. For discussions on current technology news there are gadzillions of sites in the net. What made Slashdot truly unique and fun to read were the trolls, and they have been effectively eliminated.
The troll-elimination effort has gone way too far, it's harming the technical discussions. Last week I tried to post a small snippet of perl code in my comment. It was rejected by the lameness filter, because it looked "too much like ascii art"!!!... Well, for mature and well-balanced technical discussions, posting the occasional five-line perl code is invaluable. If you don't want to see the occasional ascii-goatse, you *do* have some ways to protect your sensitive retinas, did you know that, editors?
Well, yes, I agree that the "flamebait" moderation hurts Slashdot. So, here is my own constructive proposal: when giving mod points, let only *ONE* point be negative. You get five points, you must mod four posts up, and only one post down. So you must choose really carefully which one is a troll or flamebait or overrated or redundant. It's not only the excess of flamebait moderations that's bringing Slashdot down, it's a general excess of negative moderation.
Re:since when are programs ran when they're not us (Score:2, Informative)
Re:Not a good idea (Score:3, Informative)
So, which would you rather have as a desktop - Windows 3.0 or last year's linux distro? More relevantly, who is going to be further ahead in the NEXT 10 years?
Microsoft has lost momentum, and is now reduced to trying to play catch-up in terms of features.
In terms of ease of install, and ease of maintenance, and ease of updating, linux distros win (just did an upgrade at the office from SuSE 9.1 to 9.2 Friday - almost 6 gigs of software brought up-to-date painlessly ... impossible with Windows where every service pack and update is feared for what it will break).
The problems with Windows are three-fold
- the core has too much crud that was grafted into it that seemed like a good idea at the time (performance, etc) but shows its age
... there's some really OLD stuff in there that can't be rewritten without breaking too many other things. Their designers have admitted as much. Look at the way that numerous Microsoft products use undisclosed windows api calls, then they can't modify/fix something related because those calls, while not part of the public interface, have to be maintained because removing them will cause their software to puke. They end up having to write more code to handle more special cases ... over and over and over
- Windows target market is to be all things to all people - Gates himself has said "I only want our fair share of the market - and that fair share is 100%". In trying to be everything, it is neither fish nor fowl.
- Reputation - Microsoft and Windows both have a terrible reputation, in terms of public relations (convicted monopolist, all-grasping with their "passport", etc...) and technology (everyone knows what the "Blue Screen of Death" means).
Look at how long its' taken Windows to get to where it is today (21 years + 3 in development = 24 years). Linux is half that. Wanna bet that 10 years from now Windows is still on almost every desktop? I doubt it.Linux, on the other hand, has no undisclosed api.
Linux, on the other hand, is a kernel. It's agnostic in the sense that it is up to YOU as a user/developer/distributor to do what you want with it.
The BSDs and Linux have a much better reputation. People don't use Windows because they want to, but because they have to. But that's changing.
Re:Windows clusters don't make sense (Score:3, Informative)
ptree - pslist [sysinternals.com]
w - psloggedon [sysinternals.com]
ls -al - dir
finger - finger (standard)
unzip - expand (standard, for CABs), cygwin unzip, rar
mount - (automatic), fsutil, linkd (from resource kit)
make - make (comes with SDK)
grep - find (standard)
piping with | > < are the same
perl [perl.com]
cygwin [cygwin.com] for other UNIX processing utils.