Slashdot Mirror
EDS: Linux is Insecure, Unscalable
daria42 writes "Large enterprises should not use Linux because it is not secure enough, has scalability problems and could fork into many different flavours, according to the Agility Alliance, which includes IT heavyweights EDS, Oracle, Cisco, Microsoft, Sun, Dell and EMC."
Obviously, I thought the study was biased, looking at the list of supporting companies. But then I RTFA:
The alliance comprises a group of IT hardware and software firms that have combined their expertise and products to help EDS create 'best of breed' solutions and compete with the likes of IBM Global Services and Hewlett-Packard for the most lucrative government and enterprise contracts.
Well, if Microsoft wants a lucrative government contract, clearly the organization that is supporting this move is going to decry the competition to push its own agenda.
Why do people even listen to these organizations? I suppose you know their bias from the outset, rather than having to 'read between the lines' of other organizations.
"There's no success like failure, and failure's no success at all."
- Bob Dylan
From TFH:
From TFA:
Fuji Xerox = Oracle?
FTA:
"according to the Agility Alliance, which includes IT heavyweights EDS, Fuji Xerox, Cisco, Microsoft, Sun, Dell and EMC."
Where did the inclusion of Oracle in the post come from? It is not currently mentioned in the linked article, in either the quote or anywhere else
the same EDS that charged billions for systems that don't work, and they still get more contracts ?
in todays world incompetance is rewarded and responsibility has all but evaporated in buisness
get rich and fuck over your friends, family,community and society in general cos that 30,000 sqft house is more important
perhaps a sniper rifle and vigilanties would put management and their families into line
Renault: "I'm shocked, Rick, to find gambling going on here."
Concierge: "Your winnings, sir."
Renault: "Oh... Thank you."
Mit der Dummheit kämpfen Götter selbst vergebens.
Back in 1996 EDS declared IE to be the "standard" browser for use on all internal machines. When those of us who were using Sun boxes asked "What about us?", the reply was "We have Sun users?"
DMCA - Chilling free speech since 1998.
There's nothing in Linux except the kernel.
500GB of disk, 5TB of transfer, $5.95/mo
How elegant. Linux is the kernel, so claiming that Sun forked "everything in Linux but the kernel" reduces to "Sun forked everything in Linux but Linux" and thus "Sun forked nothing." This is entirely true, but it still manages to perpetuate the myth that Sun is damaging Linux. Simply masterful. I think you have a future, perhaps in the PR department of a major political party.
The Linux kernel does have quite a few forks. They're just not big, seperate forks. Their work gets routinely folded back into Linus' fork. There's the personal forks like Alan Cox's -ac patch, Andrew Mortons' -mm patch, etc. Many architectures and sub-projects also maintain their own forks. So forks aren't bad on their own. Open Source licenses allow the different forks to share their work. It's just that the big commercial entities like to keep reminding people of the devastating Unix fork. To their commercial mindset it's the only type of fork they can imagine. And they're obviously trying to slow the commoditization of software through adoption of Open Source.
Nice way of completely missing the point.
KDE and Gnome have nothing to do with forks. They're completely different things, independently developed, and which for the most part share no code. You can't just merge them because the architecture is different.
Having several different libraries that implement widgets have nothing to do with forking. And at least Linux has only two big ones. I rarely use Windows anymore, but each time I do I'm amazed at the non-standard look of every damned application. I mean, for some bizarre reason every firewall, antivirus, IM program, office suite, etc. has to have its own widgets, and MS applications aren't an exception.
A fork is a division in the development of a program. For instance, what happened with XFree. It was stagnating, so a group of developers decided to take the current tree, and work on it separately. Result is that we now have an actually active development in Xorg. I fail to see anything bad about it.
http://www.eweek.com/article2/0,1759,1732672,00.as p
Most of the desktop computers in the UK's Department for Work and Pensions were paralyzed for four days on Monday, when a failed upgrade took them offline. The outage, covering 75 percent to 80 percent of the DWP's 80,000 PCs, is one of the largest in the UK government's not entirely impressive IT history.
And possibly one of the most costly. According to staff reports, the outage occurred on Monday afternoon, disconnecting staff e-mail, benefits processing, and Internet and intranet connectivity. According to one, a limited network upgrade from Windows 2000 to Windows XP was taking place, but instead of this taking place on only a small number of the target machines, all the clients connected to the network received a partial, but fatal, "upgrade."
Another source says that the DWP was trialing Windows XP on a small number ("about seven") of machines. "EDS was going to apply a patch to these. Unfortunately the request was made to apply it live and it was rolled out across the estate, which hit around 80 percent of the Win2K desktops. This patch caused the desktops to BSOD and made recovery rather tricky as they couldn't boot to pick any further patches or recalls. I gather that [Microsoft Corp.] consultants have been flown in from the U.S. to clear up the mess." EDS is also thought to be flying in fire brigades.
Jason Lotito
I'm pretty sure they're using Linux in the context of GNU/Linux in this case. As in Linux distributions, not specifically and entirely the kernel.
"You're older than you've ever been, and now you're even older."
Oracle is NOT mentioned in the actual article. Oracle is part of all this in some fantasy world, called Slashdot.
Why do people mention forking as a problem? If a new version forks off and you don't like it, just don't use it! Why is this a bad thing?
Because 70 versions of something that work 70 different ways mean that it is more difficult to support for network staff and software vendors.
NOT mentioned in the actual article. Remember? You're reading Slashdot...
http://grids.itmanagersjournal.com/article.pl?s
"More than half of the [world's] fastest supercomputers -- which recently might be more accurately described as super clusters that are assemblies of many lower-power processors -- run on Linux, and Top 500 super list co-compiler and original editor Erich Strohmaier does not foresee any change in the open source operating system's dominance anytime soon."
charlie harvey's website
Just wondering...
Gentoo Linux - another day, another USE flag.
... bullshit! As well all the other major enterprises that have many thousants of deployed Linux boxen running business-critical software. These folks use Linux because 1) it's much more secure and securable than the competition, 2) it scales massively, 3) they can have their own fork (e.g. apply security patches, performance changes, etc. to the current production kernel version on their schedule, not some vendor's). Isn't it ironic how some of the uses of having your own "fork" improve scalability and security. 8-)
Let's not forget that it's far cheaper than the proprietary competition even for all of those benefits.
Also, we are somewhat cautious about what happened with Unix - it splintered into eight applications -- until McNealy (Scott McNealy, chief executive of Sun) finally announced he won the battle and had the one surviving Unix out there.
Interesting that EDS shares SUN's view of what Unix versions are available. Imagine how surprised HP, IBM, and even SCO will be to learn that SUN has the one surviving Unix, considering:
The OS registered as compliant with the UNIX 03 specification is: AIX
Other "surviving" Unixes that are registered by the Open Group include, well, look for yourself... http://www.opengroup.org/openbrand/register/catalo g.htm
I disagree...
...
According to a friend of mine, Fedora Core already comes with SElinux.
People who take any pride in using their boxes properly tend to use SElinux or GRsecurity already.
Mail-servers, web-servers, browsers, and name-servers all come in multiple-process priv-dropping forms. (Or, better: don't bother fork()ing all over the place, just setuid() where you need to, to isolate modules of code in-process, for speed.) We have far more than just priv-dropping protections up our sleeve, too: propolice patches to GCC, kernel patches for virtualization (ctx, xen, UML),
So quit pontificating and apply pressure on your favourite distribution purveyor(s) to include these things by default and get out and educate the mass of people who'll only turn GNU/Linux into the next Windoze yourself!
~Tim
--
Rushing on down to the circle of the turn
Running a whole pile of mostly independant linux webservers and appservers isn't a demonstration of scalability.
Big Iron:
BigTux Shows Linux Scales To 64-Way
My current test system has 16 CPUS:
zeus0:~ # tail -15 /proc/cpuinfo
processor : 15
vendor : GenuineIntel
arch : IA-64
family : Itanium 2
(yes, it is Itanium!! Anyone got a 16-way Opteron box? Anyone? Buhler? I thought not...)
And, of course, we all know about Linux clustering:
Beowulf Clusters
Single System Image Clusters for Linux
Ignoring the oddity of Oracle being in that group, none of the rest of the members actually make a scaleable Linux box, just ones that compete with them. The slant is obvious.
- Necron69
It's true, Linux isn't really scalable.
And that's exactly why it's being used in several top SuperComputers/Clusters, including (but not limited to) NCSA's Tungsten, IBM BlueGene/L, LLNL's Thunder, BSC's MareNostrum and NASA/Ames' Columbia.
Count them, that's 4 out of the 5 fastest [publicly known] SuperComputing clusters.
Are you really that unfamiliar with different ways to quote words? Here are a couple of common ones that apply in this instance. I will show them to you so you don't get confused again.
*word* is equivalent to word. On a system with no facility to produce fonts of differing weights, quoting the word with asterisks shows emphasis in place of what would otherwise be bold text.
'word' is equivalent to <sarcasm>word</sarcasm>. In a conversation where facial expressions and inflection are not available, the <sarcasm> tag or single quotes are often used to convey a sarcastic tone.
</rant>Now you should be able to differentiate between the two. Next!
"Murphy was an optimist" - O'Toole's commentary on Murphy's Law
I was just illustrating that having parallel platforms stagnates progress.
Rubbish - choice is a Good Thing. I like being able to choose my Window manager and have the WM's developers make all the decisions for me. I was using Gnome for a while (I find KDE completely unintuitive) and I got sick of the way the Gnome project was heading so I switched to Enlightenment and am very happy - you can't do that under Windows which it's single shell (yes, I know there are 3rd party shells available for windows but if you bring them into it then you just destroyed your own arguement).
Note: I have no objection to people using Gnome / KDE / whatever they like - if it works for you, fine, but taking away the user's choice under the false claim that choice causes stagnation is rubbish.
I'm sure you can explain how the lack of choice in web browsers has prevented the stagnation in Internet Explorer can't you?
and can copy-paste damn near anything between each other
I have had no problems with copy and paste under X for years. Windows and the various WMs under X are _different_ - this doesn't make Windows _better_. I've not used Windows seriously since Win98 and every time I have to do something with Windows I realise just how much easier and more intuitive Linux is.
http://blog.nexusuk.org
As a former EDS employee you've got no grounds to say anything.
Whether you like to hear it or not, the parent poster's opinion is shared by many. EDS is THE worst IT company I've ever had to deal with and that goes from the top to the bottom. Actually it's hard to tell how competent their staff really is because they're never around long enough to find out. It was rare to see the same faces at technical meetings. Any employees that do actually try to help the customer are generally the first to go.
Even a blind pig gets an acorn once in a while so by sheer luck EDS is bound to employ a handfull of keepers, but just generally they suck. I wouldn't hire EDS to run CAT5 to my dog house.
That's our life, the big wheel of shit. - The Fat Man, Blue Tango Salvage
Good article from all the way back in 2004 regarding where this is actually pointed. http://www.crn.com/showArticle.jhtml?articleID=510 00391&flatPage=true
e x.html. Guess not.
m l
s p
Would Sun rather see Linux go away? Sure, but they also believe in it enough to sell it. http://www.sun.com/servers/entry/v20z/index.jsp
These are quotes directly from they guy heading up EDS's strategic alliances. Not from members of the strategic alliance - has anyone asked Ellison if he thinks Linux is insecure, prone to unfriendly forking? Guess not. http://www.oracle.com/events/unbreakablelinux/ind
Cisco? Well lets see they have linux running on some of their hardware, and apparently its good enough for their engineers to run http://www.nwfusion.com/news/2005/0216cislinux.ht
So lets round out the list...
EMC - http://www.emc.com/products/systems/linux/index.j
Dell - http://linux.dell.com/
Microsoft - http://www.mslinux.org/ Err, umm - ok maybe not.
http://windows.scares.us
That EDS is a "Microsoft Partner" (Windows whores) and they haven't even managed to impelement single-sign-on for all the apps on their corporate network yet. And Windows has really taken them a long way with that US Navy contract. Maybe if they'd gone with a UNIX based solution, they wouldn't be years late and unpaid for that thing...
I'm trying to teach myself to set people on fire with my mind... Is it hot in here?
They say linux doesn't scale well. SGI has Linux systems with 256 cpus in a node. http://www.sgi.com/products/servers/altix/ Microsoft is only now getting a cluster version of their OS http://developers.slashdot.org/article.pl?sid=05/0 3/04/2134229&tid=201&tid=231&tid=156
would highly resent having to replace all it's fine work on NMCI with a new technology. Even if it would be an improvement.
-- Improve Windows - Buy a Mac!
It's a bit like saying that florescent lights are scalable because you can put thousands of individual lights within a building, or that IBM laptops are scalable because you can purchase them in units of 1000 running MS Windows.
In a sense you're right. One flourescent light is not scalable. But the technology itself is: it's cheap, it's reliable, it's easy to maintain, and adding more bulbs gets you more light.
The same goes for the laptop analogy. In this case, the laptop is a means for the worker to do more work in a given time frame. Clearly if you had 1000 workers but only one really big laptop you'd be in trouble. Workers can "scale" if they managers can provide enough computers (yes, I know large teams often have diminishing returns).
It may not be mentioned in the article, but Oracle are mentioned on the Ability Alliance's membership page. The Slashdot summary is completely correct.
Slashdot monitor for your Mozilla sidebar or Active Desktop.
And this was deemed "informative" ? For crying out loud: the author hasn't even bothered to check that his point is moot. Even if not mentioned in the article, Oracle is still a member of EDS.
If you check the Alliance's web site (http://www.eds.com/services/alliances/agility/) you will see that Oracle is indeed a member of that group. As is SAP and Siebel, but, they aren't mentioned in the article either.
The poster is correct.
I work for EDS (hence the anonymous post). Have you had to deal with a customer who changes requirements on a whim and then proceeds to threaten you with severe penalties if the system will be late? The net result is poorly tested systems that generally don't work.
Hi,
I am the Linux systems specialist at Welchs. yes...
the company that brings you that awesome purple grape juice. We are in the middle of an ambitious
ERP project migrating from a Mainframe/AS400 to a
Linux / Oracle RAC solution. We are running Linux Clusters and Oracle RAC. We've got connectivity to our EMC SAN. I've got a development, QA, and Production landscape. Here's our PROD landscape layout, 2 Oracle Database RAC Nodes, 4 application servers loadbalanced with F5 load balancers, and we seperated out the concurrent managers and have
2 concurrent managers (oracle cluster manager controls those nodes). Linux is highly scalable.
I'm also implementing Linux HA to deliver some
highly available filesystems to our cluster.
Announcing KDE 3.4
That is what the web is all about.
Well, what I meant was that you were either smart enough or fortunate enough to get out before things went downhill. In either aspect you're not part of the problem now. Your information is out of date unless you just left last week.
EDS hasn't always been bad to work with, just in the last few years. Whether it's GM's influence or what, I absolutely hate working with them. All they want to do is count milestones, doesn't matter what minimal, sloppy work it takes to get there. They'll go for the low-hanging fruit to claim success, then try to go back and make it work. Agh, I f'ing hate that. And you have to audit their billing constantly for over-charges. I've worked with Dell prof. services, IBM, HP but the only one I can't get along with is EDS.
And it just figures within minutes after I post that I'd run into two EDS employees who were very nice and seemed fairly geek competent.
I give them a month, maybe two. :)
That's our life, the big wheel of shit. - The Fat Man, Blue Tango Salvage
If Tyranny and Oppression come to this land,
it will be in the guise of fighting a foreign enemy. -James Madison
As someone who worked on that project (NMCI)...it is definately as defined above....oh...but they use linux for many of their file servers on the project...hmmm....I know this due to the groaning from Tier2 when they had to reset permissions on them through MS terminal services (anyone on the project ever hear of the CLI...really) EDS is contradicting themselves,and as a former employee, this is no surprise.
Nope. The IE team completely rewrote their controls to be windowless. If you don't believe me see here.
Here is a slightly more elaborate system, which minimizes downtime.
- Install the new drive in the server alongside the existing one.
- Partition it appropriately. I will describe how to move one file system to the new drive; any others can be done similarly.
- Mount the new partition somewhere temporary, such as
/mnt.
- Copy all files to the new drive: cp -ax <old mount point>/.
/mnt
- Shut down all processes using the old mount point. (Single-user mode with init 1 is probably the easiest way to do this.)
- Update the copy to reflect any changes made while it was being copied: rsync -aHx --delete <old mount point>/
/mnt
- Unmount the old mount point, and
/mnt.
- Update
/etc/fstab to install the new partition in place of the old one.
- Mount the new file system in place of the old one.
- Restart all tasks, daemons, etc.
With various LVM systems, or something fancy like Sun's ZFS, you can do it all on-line, but that's the basic technique which works for anything. There are, of course, many other ways to copy the contents of a partition around; cp amd rsync are just two. There are two things I know about which could need special handling, because they depend on the inode number of the copied file: qmail spool directories and nethack saved games. Both packages include utilities to fix things up after changes such as the above. Changing the root partition requires a slight modification to the final steps, using the pivot_root(8) utility. (Or you can just reboot the machine.)Actually.. I think the NMCI users are being very kind by only calling it a royal clusterfuck
The fact of the matter is that the EDS loadout on NMCI machines is outragously slow.. prone to crashes... and generally about useless..
My command had 500 plus NMCI machines on our legacy network while we were waiting on EDS to build our infrastructure... these machines were provided by EDS but we used our W2k load outs on them.. they ran flawlessly...
Once EDS finished the buildout of the infrastructure they wiped these exact same systems and loaded thier loadout (W2k, Office2k... and a few other apps we need)... the performance went into the crapper! The machines crash ALOT... many a day I come into work and have 1 or 2 computers with a blue screen due to some patch they pushed in the night.
They state that they don't lock out the USB or seriel ports and that as long as the item we're adding has W2k drivers native on the system (ie things like USB thumbdrives, mice, keyboards, etc) we should be able to plug them in and have them work... they don't! You have to get an admin to come log in locally.
The amount of money my command pays for this service each year is 3 times our old budget with less computers, less service, slower tech response times and generally disgruntled users.