-
Website
http://www.scobleizer.com/ -
Original page
http://scobleizer.com/2006/06/06/windows-more-reliable-than-redhat-linux-yes-says-yankee-group/ -
Subscribe
All Comments -
Community
-
-
Top Commenters
-
Popular Threads
People are always looking for some panacea and they will never find it. People will always be the weak link where software is concerned. Use Linux where Linux excels and use MS where it excels.
Why have a 64bit machine if you can't run a 64 bit OS.
Robert, we can see what you are trying to say here, but it doesn't work. If something is stable and you add to it, that doesn't guarantee the result is stable. In fact, you are likely to mess things up.
Example:
http://www.businessweek.com/the_thread/techbeat...
Sorry, Scoble, but I wouldn't trust anything that came out of that organization.
Every salesman will through a freebie in when they're hoping to get more sales soon!
I'm not saying if they got it right or not, but I am saying that as far as I'm concerned, they're a little 'tainted'.
Bad analogy time: you're a supermodel, paid to make a particular makeup sell. Your contract expires, but you know that the new product is just around the corner and will need some talking up, too. Are you going to just bad mouth the old product, just because you're no longer paid to advertise it?
Hey: like I said, Devil's advocate.
I guess it's good that they've got the "start" part accomplished, but I think people are more interested in when the waiting and delaying will end.
It's pitiful that they have to describe their report as "independent, non-sponsored".
A reputation is a terrible thing to lose.
According to the Yankee Group's annual server reliability survey, only Unix-based operating systems such as HP-UX and Sun Solaris 10 beat Windows on uptime
So it's not like Windows was #1 here.
Furthermore, what does that 20% mean?
Additional key results for the independent, non-sponsored Yankee Group 2006 Global Server Reliability Survey show that on average, individual corporate Linux, Windows and Unix servers experience three to five failures per server per year, resulting in 10.0 to 19.5 hours of annual downtime for each server.
So, considering Linux is near 19.5 hours and Windows near ten, that's a difference of 9.5 hours. In a year.
If that's the only server you have running, that's bad. If you follow best practices with regard to redundancy, then that's how long it takes to reimage and restore changed data with no loss of productivity whatsoever.
That's hardly a major difference, especially considering that Solaris and other high-end Unixes beat Windows. (By how much, we don't know. Funny how that got left out.)
Maybe Vista should have been built on top of Solaris 10 instead.
Of course, since I wasn't able to find the actual survey data, we're stuck with what the Yankee Group chooses to tell us. That alone makes any results, good or bad essentially meaningless, since we can't study raw data and methodology.
Real people use OpenBSD. OpenBSD will kick Win2k3's ass. I run an OpenBSD 3.9 server, and I just migrated from Windows Server 2003 Enterprise.
For Windows to have 20% more uptime than Red Hat, Red Hat would have to have an uptime of less than 85% and Windows 100%. A few pico-seconds of thought would show they got the numbers wrong.
(they may have meant 20% less down time but if you can make a stupid mistake like that how could we believe them)
Back off man! You're in my space.
Anyhow, eveyone knows Didio is a shill.
Thanks, man, you've made my day! :-)
As someone who has significant firsthand experience overseeing administration of Linux (not Redhat, tho) servers and Windows 2003 servers (exchange, yuck), I can conclusively say that finding is complete bullshit.
One machine is still running on the same Slackware install for two years serving web apps, getting updates with no downtime and getting a scheduled restart every 110 days.
The exchange server running Windows 2003 is restarted about once a week and has already been wiped twice with a complete OS reinstall in the last 6 months because it was compromised even within the corporate firewall.
Guess what's built starting with Windows 2003 Server's code base? That's right, Windows Vista.
I doubt that's true, but an operating system used for client and server faces completely different points of failure. While a server runs a limited set of applications pretty much all the time, a desktop client goes through a lot more application starts and stops and as a result is more prone to developing memory leaks which cause crashes.
MS Windows excels for video gaming, but I still think that's more a function of marketshare than any technical superiority.
When it's time for content creation, I use Macs. When it's time to serve or reach out to customers, it's time for Linux.
Lets see the same tests with THESE configurations:
Windows Server 2003 vs OpenBSD 3.9
Windows Server 2003 vs Gentoo 2600.0
Windows Server 2003 vs Slackware 10.2
Windows Server 2003 vs FreeBSD 6.1
Comeon Yankee Group, or are you too scared because you know that if you tested Windows against a better *nix system, that Windows would lose. Thought so.
btw, windows 2003 is sure a lot stable.
God knows what idiots might type when you're away, like ANOTHER bogus 'MS is so great, linux totally sucks' report.
Have you ever tried a modern linux distro?
It seems she concludes how well managed a system is makes more difference to up time than choice of OS. What next ? A report saying rain is wet ? HP-UX and Solaris are more reliable. Well with a fixed set of hardware support (a luxury neither Linux nor Microsoft enjoy), support contracts as the norm (ditto), and a much narrower range of applications, you'd expect that. If you compare Solaris boxes which run only Oracle, with a support contact, and dedicated IT staff, with a mixture of all Linux or all Windows environments what result would you expect ?
It seems the people who have to find something intereting about the report in order to get customers for Yankee strugled - hence the "Windows better than Linux" line - which the trolls find so hard to accept. (e.g. #30), the report isn't "MS is so great Linux totally suck" - it is "Linux and Windows are about as reliable as each other, but Linux takes a little longer to fix". A total non story "OS reliablity: everything much the same" doesn't get much attention.
I was getting effectively 100% uptime from Novell Netware 2.0a back in 1989 (power supply to our building could only manange about 99.7%) - how did I manage it on software which wasn't all that good ? The installation meant you couldn't stray far from best practice, and once installed you left well alone. If that were true of todays systems (Linux or Windows) - we'd have better up time.
"Custom SuSE Linux delivers the highest reliability and fewest minutes -- about 430 minutes of outage per server, per year."
I don't find any RedHat based linux distro (the kind that can use .rpm's) very reliable unless you compile everythign yourself.
Debian is one of the most rock solid distros out there. Most hosting companies use Debian because of this.
I don't understand WHY people use RedHat for a server, to me it just doesn't make sense. I can uderstand if they are new to linux or are going to be using RedHat for a desktop.
Now, it's still 1,000,000% better than Windows 98. I hardly ever ctl-alt-del anymore. I don't think XP is better than 2000. If ctl-alt-del only appeared as quickly as it did in Windows 95, it would be perfect. Because now when something goes wrong, it really goes wrong, and I need a hardware reboot.
Now, I don't mean to be critical to MS. XP sp2 is a great product. Windows usually runs great on older hardware (up until Vista, they did better at this than anyone). They are too conservative in a number of areas, but I can still play Starcraft on XP and my company still runs Dos software (thankfully, I personally don't, I log in to Linux). But the difference between Vista and Server 2003 has to be lightyears, or your marketing dep't has some explaining to do.
How odd that the X360 isn't truly backwards compatible?
Like others have commented, you need to find an analyst with some credibility if you don't want to be laughed at for repeating this stuff.
Go ahead and submit it to slashdot! They'll love it!
http://linux.slashdot.org/article.pl?sid=06/06/...
So if Windows had 20% more uptime than Linux annually, then that means the Linux system was down for *60 days straight*? Was an MCSE operating the Linux system or something?
Scoble--who are you kidding with the "not sponsored by Microsoft" line? The Yankee Group and the Laura DiDio (people conducted the study) have a history of writing shill articles for Microsoft.
Here's an article written a year ago about the same people who conducted this study:
http://www.businessweek.com/the_thread/techbeat...
But Scoble blindly walks in posting whatever moves...armed with not even a pea-shooter, nor a single-cell of brain-power, unaware of the history about it all.
How can they get away with such blindly incorrect stats. They should retract and publish the stats the mathematicians say are accurate!
I suppose this is off topic, so feel free to delete it.
As to iBook or iPod or Playstation, what does that have to do with the reliability of Windows? Well, the iBook might, but Patrick has a Windows machine too and now he wants an Intel machine so he can run Windows on that (Windows has a lot more games than his Mac does).
Playstation 3? Nah. We already have an Xbox 360. Don't see what I'd get by spending another $600.
A Macbook Pro then? ;-)
i don't know if it's better that redhat but windows 2003 it's surely stable (and a lot better than xp)