+ Reply to Thread
Results 1 to 2 of 2

External library not being pinned in cache...

  1. External library not being pinned in cache...

    Hi, all.

    Running 9i (don't have a version number) on Solaris.

    A stored procedure is calling an external C library.

    I'm looking to confirm whether or not this C library is being held in
    memory after it's called. The user is theorizing that it is not, and
    that it's being retrieved from disk each time it's called.

    I'm not sure whether I can expect that kind of external routine to
    stay pinned in the library cache or not. And if not, what the best way
    to ensure that it *does* stay pinned...

    Comments appreciated...

    BD

  2. Re: External library not being pinned in cache...

    On Aug 15, 2:53*pm, BD wrote:
    > Hi, all.
    >
    > Running 9i (don't have a version number) on Solaris.
    >
    > A stored procedure is calling an external C library.
    >
    > I'm looking to confirm whether or not this C library is being held in
    > memory after it's called. The user is theorizing that it is not, and
    > that it's being retrieved from disk each time it's called.
    >
    > I'm not sure whether I can expect that kind of external routine to
    > stay pinned in the library cache or not. And if not, what the best way
    > to ensure that it *does* stay pinned...
    >
    > Comments appreciated...
    >
    > BD


    I'm thinking this is a feature, see metalink Note:302485.1 . The
    feature being:

    "The function execution occurs in a seperate address space other than
    the Oracle servers address space , and this totally isolates the
    server from the effects of external procedure execution."

    I'm guessing you could pin the package that calls the function, but
    that doesn't address your user's worry, since the library is outside
    of Oracle. On the other hand, I think buffering on the Solaris side
    might make it moot, so you should be careful with how you interpret
    system statistics. "See your System Administrator."

    jg
    --
    @home.com is bogus.
    BNAFH: http://www.infoworld.com/article/08/...ockout_1..html


+ Reply to Thread