Slashdot Mirror
Three Enterprise Operating Systems Compared
Anonymous Coward writes "Finally, a much awaited review of enterprise OSes. The guys from NW Test Alliance pitted
Red Hat, UnitedLinux, and Windows against each other and rated them on several rubrics. Red Hat won by a slight margin on the basis of its high hardware compatibility and strong security integration."
Read the article. There's a graph with some stats on Windows vs the two Linux distros, but it's not a comparison between all three - only between the two Linux distros. The last page makes it pretty clear when they only rate the two Linux distros, and Red Hat wins that comparison.
This is *not* a long-awaited comparison between Windows and Linux. It's not even a long-awaited comparison between Linux distros - the whole article spans a whopping three pages, and it's woefully incomplete.
What's your damage, Heather?
According to note at the bottom of the article, the results for Windows Server 2003 came from a previous test (I didn't bother to try and search for it, asthey didn't provide a direct link). It would seem that the comparison would be more valid if the tests were all done at the same time, or at least on the same hardware and have some statement to that effect.
I'm not trying to knock on the test, but just pointing out that even smal changes in hardware components or settings can make a big difference.
Otherwise, it looks like a good and thourough test.
One never knows whether a journalist/reviewer/linux-advocate really understands what an "enterprise"-ready OS is. For the purpose of this post, I'm not arguing whether Linux is or isn't one. But I had to laugh after seeing a chart showing "Successful transactions per second" and doublechecking their footnoted definition of transactions.
OLTP? Database? TPC-C? No. A transaction was downloading 20 4k-byte files.
--LP
I think they can justify it. I mean, when you're used to paying almost $3,800 for Windows Server 2003 Enterprise Edition with 25 CALs, $2,500 (and no CALs) sounds pretty good!
Q: "Why do sound techs say 'check 1, 2'?"
A: "Cause if they could count any higher they'd be lighting techs."
Anyway, I'd like to see a comparison for the major players of the real enterprise OS market: z/OS, OpenVMS, Solaris, AIX, Tru64, and HP/UX.
This is an interesting test that I haven't seen done before. Interesting to note that Suse took much longer to reply to the emails, although the article doesn't mention if the Suse support people are located in Germany, and if the time zone difference could be the cause. Red Hat's more detailed responses sounds like a plus, though. Although I would like to have seen the actual questions and responses. Anyway, this sort of thing is important for a company like mine, where we use Linux, but can't (or won't) afford 24/7 support (I should mention that Linux isn't a primary platform here, we do have 24/7 vendor support for our mission critical systems). So getting a quick response on emails is a big selling feature.
They should have at least mentioned in the article that Windows was tested on 100Mb ethernet while the Linux systems were tested on GB. They did provide links to the test configurations, but this is a major fact to leave out, at least when you are testing network performance.