OpenBSD Journal

Computer Business presents a nice showcase of 3.6's features.

Contributed by grey on from the security, SMP, and free? OpenBSD gets more big iron credit dept.

In the article SMP-capable OpenBSD 3.6 set for November we see some Computer Business Review Online give some nice coverage of the upcoming features for OpenBSD 3.6 and how it compares favourably to other commercial unix vendor offerings.

Indeed, the push to get SMP going in OpenBSD as a priority feature for 3.6 should make a lot sense for those who haven't been paying attention to the direction the cpu industry is taking in the coming months.

(Comments are closed)

Comments

  1. By SH (213.15.68.89) on

    Nice article, but reading about the new 3.6 features the article mentioned, it appears that the author used an older version of http://www.openbsd.org/36.html when writing it.

    /SH

  2. By Stephen Paskaluk (129.128.138.50) sap@66h.42h.de on

    Is it just me or does this article show a lot of poor research and editing. They refer to HP's "AIX for its PA-RISC architecture" claim that the ports of HP-UX and AIX to Itanium mean that they could be ported to opteron as well (they could be, but it really has nothing at all to do with itanium). They claim OpenBSD's support for sparc and sparc64 is due to it being a "close cousin" of Solaris. Also, I'm not sure if the amd64 port will work on the 64 bit intel xeon machines as the article claims (though the standard x86 port works fine).

    There are other small mistakes too, quite frankly I'm shocked that this article was actually published as is. All in all I think it was a pretty poorly written article, though it'll probably provide good publicity for OpenBSD.

    Comments

    1. By Brad (216.138.200.42) brad at comstyle dot com on

      amd64.html clearly states... "It also runs on the upcoming Intel ia32e processors which contain cloned support for the AMD64 extensions".

      Comments

      1. By Stephen Paskaluk (129.128.138.50) sap@66h.42h.de on

        amd64.html clearly states... "It also runs on the upcoming Intel ia32e processors which contain cloned support for the AMD64 extensions".

        That's my bad, I was remembering an earlier statement when they first got the Intel kit to test which referred to them running the i386 port on it. In my defense I did acknowledge that I wasn't sure about the status of that :)

    2. By Sean Brown (68.147.170.205) on

      Do you mean where it says, "...brought HP-UX functionality on Itanium on par with AIX for its PA-RISC architecture." Here, let me reword it for you because you seem to have a problem with it. It doesn't say AIX on PA-RISC, it says HP has brought HP-UX on PA-RISC on par with IBM's AIX on Itanium. Its talking about a performance comparison.

      Comments

      1. By Stephen Paskaluk (129.128.138.50) sap@66h.42h.de on

        Do you mean where it says, "...brought HP-UX functionality on Itanium on par with AIX for its PA-RISC architecture." Here, let me reword it for you because you seem to have a problem with it. It doesn't say AIX on PA-RISC, it says HP has brought HP-UX on PA-RISC on par with IBM's AIX on Itanium. Its talking about a performance comparison.

        I disagree with your reading, since your interpretation ignores the obvious HP-UX on Itanium reference, and HP-UX is very well established on PA-RISC. They're talking about the level of support for the Itanium in AIX and HP-UX, clearly not a performance comparison. I think it was a simple typo that should have had HP-UX instead of AIX, but for some reason you think an out of place comparison that isn't supported by the text of the article is more likely than one more typo.

    3. By Otto Moerbeek (213.84.84.111) otto@drijf.net on http://www.drijf.net

      There are other small mistakes too, quite frankly I'm shocked that this article was actually published as is.

      Oh man, you are easily shocked. This is an online, business publication. What's important is the general message. Like other posters already mentioned, the errors you are seeing are not really errors. And after all, Solaris is BSD derived.

      But mickey might be shocked because it says hppa64 is on the wane....

      Comments

      1. By Marco Peereboom (67.64.89.177) slas@peereboom.us on Marco Peereboom

        Yeah Mickey! Get to work :-) I want to fire up those B2000 that I have :-)

      2. By Stephen Paskaluk (129.128.138.50) sap@66h.42h.de on

        Oh man, you are easily shocked. This is an online, business publication. What's important is the general message. Like other posters already mentioned, the errors you are seeing are not really errors. And after all, Solaris is BSD derived.

        I know Solaris is BSD derived, but that really doesn't impact the sparc and sparc64 ports as near as I can tell. Solaris (unlike the earlier versions of SunOS IIRC) is also very much SysV, and I don't think calling Solaris and OpenBSD close cousins is a very accurate comparison, not more than any other Unix-like systems can be called close cousins.

        I suppose you think my criticism is unfounded and that it was a well written article then? The general message is important, but details shouldn't be ignored. If people making decisions feel that the article is poorly researched they're less likely to accept the general message.

  3. By jtorin (194.103.189.24) on

    What I feel is a shame is that none of OpenBSDs security features are mentioned. SMP in all its glory but to be frank, most other unices in competition has had it for many years... So it's not that big news for the rest of the world.

    I consider OpenBSD a good choice not because of performance, but because of security and stability. And if you read Theos slides you realise how many techniques, large and small, that have gone into OpenBSD at this stage.

  4. By Eduardo Alvarenga (66.110.114.5) eduardo at thrx dot org on http://www.thrx.org

    "The 3.6 release will include OpenSSH 3.9 for system administrators, as well as the normal slew of open source Unix tools that have been made famous by Linux, including XFree86 4.4.0, Gcc 2.95.3, Perl 5.85, Apache 1.3.29 and Apache 2.8.16...." He is really nuts :)

  5. By nikns (195.122.29.100) nikns@secure.lv on pazeme.lv

    Sorry for my schizo, BUT does this smp support supports my intel cpu with hiperthreaing technology? (i have only one cpu). Intel says, that HTT gives 20% speedup. it would be nice if i could use this feature on mine openbsd boxes.

  6. By Kevin Kadow (163.192.21.44) on

    I just installed a 3.6 snapshot on a dual processor Dell 1750, works great, SMP kernel boots without a hitch, performance is outstanding.

    I'm actually quite suprised at the push to incorporate the very recent SMP changes in a -RELEASE version, it almost seems to go against the OpenBSD philosophy to make such a major change in such a short time.


    For my purposes, the performance boost will be a nice bonus, but I am slightly concerned that stability might be degraded?

    Comments

    1. By Brad (216.138.200.42) brad at comstyle dot com on

      SMP is NOT ready for production use. The initial work was put in and tested to ensure we did not have regressions with uniprocessor systems. SMP is a massive undertaking and there is no way we could have just dropped in SMP code overnight and expected it to go off without a hitch. What is there now is for *testing* purposes and to provide a basis for future work.

Latest Articles

Credits

Copyright © - Daniel Hartmeier. All rights reserved. Articles and comments are copyright their respective authors, submission implies license to publish on this web site. Contents of the archive prior to as well as images and HTML templates were copied from the fabulous original deadly.org with Jose's and Jim's kind permission. This journal runs as CGI with httpd(8) on OpenBSD, the source code is BSD licensed. undeadly \Un*dead"ly\, a. Not subject to death; immortal. [Obs.]