Scyld to Release Beowulf 2
grantedparole noted that Scyld has announced that they'll be releasing Beowulf 2 on tuesday. Scyld's CTO is none other then Don Becker. Presumably they'll be showing this off at ALS this week (ALS is the Atlanta Linux Showcase, and is probably the best of all the Linux shows. Since its in Atlanta, its also the only tradeshow that doesn't require me to take a connecting flight!).
Alex Bischoff
---
Alex Bischoff
HTML/CSS coder for hire
...and this time, it's personal.
Sorry - couldn't resist it.
Hacker: A criminal who breaks into computer systems
"Information wants to be paid"
Looks interesting - a good, standard-ish Beowulf distribution with knobs on would be most appealing. But, my applications require large amounts of database-type searches against huge files that change rarely. I want to be able to store that data on the local drives of my cluster rather than shunt it around the machine. I don't think that is very easy to do at the moment... It would be nice if there were accepted ways to do that. Does anyone have any advice BTW?
"Scyld Beowulf's unified approach extends to the systems installation and boot designs. The operating system on cluster slave nodes is downloaded from the front-end computer. Only a minimal boot image on either a floppy, a CD, or the cluster node's hard disk is required for each cluster node. Once booted, cluster node configuration is controlled by the front-end"
This is a great feature.. You can upgrade the Kernel or Apps on the entire cluster by simply patching the Master Node..
I think it's time to start collecting old Compaq's again. But where will I put them all?
Pontiac
If you think it's expensive to hire a professional to do the job, wait until you hire an amateur. --Red Adair
From the kernel monte page it's obvious that only i386 architecture is supported. To bad they don't say it right away on the home page.
Don Becker, while at NASA's atmopheric labs, had the vision of using many Linux boxes as a cluster before Linux was even at the 1.0 revision level. He is often creditied as being the inventor of Beowulf. He wrote over half the network drivers in Linux and Beowulf as a means to an end.
This is a boring sig
Morons!
- A.P.
--
* CmdrTaco is an idiot.
"Remember when the U.S. had a drug problem, and then we declared a War On Drugs, and now you can't buy drugs anymore?"
[Sorry about my other crap post...return in the subject textbox submits :( ]
MOSIX has transparent fork-and-forget process migration, but is a bitch to set up. It would be nice if the Beowulf-style setup (and process management tools) were available in a MOSIX like cluster. Beowulf is still for MPI apps and MOSIX is still for "normal" apps, but it will be interesting as these two products develop to see where they overlap...
It is from the classic poem
From their page
What does Scyld stand for? Is it an acronym?
-The name Scyld comes from the Beowulf legend. It also works out well as an acronym
http://www.scyld.com/FAQs.html
So if the name of the company is from the legend....follow me?
I still dig it though, and have actually read it several times on my own. (A cool modern Sci-fi adaptation is the beowulf series by Larry Niven, which include Beowulf's children and the Children of Heorot (released under a different name in the UK)).
rev
The signal-to-noise ratio has reached an all-time low with this discussion. Intelligent conversation has been replaced with immature rantings and stale jokes.
The whole site is now hopeless. The first amendment has a lot to do with it.
The Scyld Beowulf-2 distribution you can buy from LinuxCentral.com is x86 only, but we support other architectures.
The only x86 specific feature is the cool "Two Kernel Monte", a kernel module which allows you load a new kernel(!). T-K-M is useful for any Linux system, not just for Beowulf.
The Alpha AXP is supported only with custom distributions because the Alpha requires a kernel matched to the specific motherboard type. That would mean two dozen CD-ROMs instead of just one.
We previously had Sparc-32 support, but that has been dropped. Beowulf is focused on price/performance. Sparcs are expensive and slow.
PowerPC support is planned. The Beowulf-2 system is based around BProc, which requires processor specific modifications to the kernel. For instance we add a new executable type to the kernel and "VMA dump" to save an executing program to a file or network stream. So it's way more than just a recompile to support a new architecture.
--To the tune of Van Halen's "Panama"
Jump back, what's that sound?
Here it comes, case open power down.
Cold boot, running through the fs check.
Awesome parallel. Running on Intel.
Don't ya know I'm gonna 0wn SETI?
I'm gonna factor primes.
I'll find pi....
Beowulf!
Beowulf!
Beowulf!
Beowulf-ulf-ulf-ulf-ulf-ulf-ulf
Ain't nothing like it, a cruddy machine.
Take some old Pentiums and a Red Hat CD.
NIC cards, buy 'em by the crate and save.
Got some 100-base running through my bedroom.
Don't ya know I'm gonna 0wn SETI?
I'm gonna factor primes.
RC5....
Beowulf!
Beowulf!
Beowulf!
Beowulf-ulf-ulf-ulf-ulf-ulf-ulf
(spoken)
Yeah, we're running a little bit hot tonight.
I can barely see the screen from the heat comin' off of it.
I reach down, between my legs....
Pop the CD tray....
You're flustered. I'm clustered.
Like a Cray running in my closet now.
Got the boxes. Alan Cox's.
Mips a floppin', ain't no stoppin' now!
Beowulf!
Beowulf!
Beowulf!
Beowulf-ulf-ulf-ulf-ulf-ulf-ulf
-- In the future, everyone will code Perl for 15 minutes. --
MOSIX and "beowulf"-type clusters are totally different beats.
MOSIX just migrates the application to one computer, beowulf is about letting it run on many computers at the same time. To do the latter, the applications need to programmed differently - no matter how people would want it to be, someone needs to split the computation into parts which can be computed at the same time. Some operations can be done by compilers (like for loops without inter-dependecies in the data) and others (like some common linear algebra routines) can be placed in libraries, but there's not going to be an automatic system for parallellizing everything: Knowing what parts of the problem can be computed at the same time and what data is needed to do so will continue to be important for scientific apps.
There is no overlap.
I wish I had my mod points now. This has to be the single most idiotic, useless, unconstructive, narrow-minded, bigotted, stupid, load of crap I've ever read on /. (and that's up against pretty stiff competition). To quote Greg Stafford, you know very little, and what you know is wrong.
"Encyclopedia" is to "Wikipedia" what "Library" is to "Some people at a bus stop"
If you would be interested in signing up for the troll mailing list, click here.
rev