Die Aufregung war gross, als letzte Woche bekannt geworden ist, dass Oracle 10.2.0.3 nicht auf VxFS betreibbar ist. Jetzt ist ein One-Off-Patch verfügbar, aber die Schlamperei seitens Oracle scheint weiterzugehen, wie Kevin Closson schreibt. The 10.2.0.3 Patchset with VxFS Saga: An Example of Incorrectly Describing the Incorrectness:

Since the Metalink note got it wrong by stating that VxFS doesn’t support “directio” (a.k.a Direct I/O), I’ll clear it up here. As I stated in this blog entry, the true nature of the bug is that the 10.2.0.3 porting team implemented a call to the Solaris directio(3C) library routine which is a way to push Direct I/O onto a UFS file, but is not supported by VxFS. There, now, doesn’t that make more sense? Am I being a stickler? Yes, because there is a huge difference between the two following phrases:

  • attempted to use directio with vxfs
  • attempted to use directio(3C) with vxfs

Kevin impliziert damit, dass Oracle bewusst die Tatsachen verschleiert, um Propaganda gegen VxFS, und damit für ASM, zu machen, aber ich kann ihm da nicht recht folgen. Jedes technische Bulletin ist ein Kompromiss aus Details und Verständlichkeit, die sich den verfügbaren Platz streitig machen, da hat meistens keinen Platz mehr für Corporate Bullshitting…
(Und natürlich hätte ich als Engineer auch gerne mehr technische und korrekte Details in den Metalink Notes.)

4 Responses to “Patch für VxFS mit Oracle 10.2.0.3”

  1. Kevin Closson Says:

    On the contrary, I did not imply anything. I stated matter-of-factly that the Metalink note stated “directio” when it should have stated “directio(3C)”. That incorrectness is very sloppy.

  2. maol Says:

    After reading your article again I have to agree, you’re not implying it in this article. It’s probably just that from following your blog I have the impression you don’t like how Oracle is marketing ASM, which made me think you also imply that in this article.

  3. Kevin Closson Says:

    Yes, Maol, you are correct that any time I specifically blog about the topic of ASM I will have an anti-ASM point of view. I feel my indictments of ASM are clear and not implied whatsover. As I point out in my postings, ASM partially solves a problem that does not need to be solved. I think over time NFS will become more of a popular option–particularly after the release of 11g.

    I’m quite certain there was no covert anti-Veritas messaging when they produced the Metalink note using the term “directio” where they should have used “directio(3C)”. I think it was more likely that the author of that note doesn’t know the platform that well.

  4. maol Says:

    Absolutely, I see we agree now.
    I wonder though about the adoption of NFS for datafiles, as that’s still one of the bigger taboos for databases. And yes, I know that Oracle’s using NFS for >90% of their hosting environment, but I’m still reluctant to use it enterprise-wide…