Re: UTL_FILE on an RAC cluster enviroment

From: John Hurley <johnbhurley_at_sbcglobal.net>
Date: 23 Nov 2004 05:52:25 -0800
Message-ID: <d4d6f278.0411230552.59b6ded3@posting.google.com>

When you run a plsql proc that uses utl_file, it runs on "one" of your databases instances.

You can check instance_name from gv_$instance if that matters to you (to see what instance your proc is executing on).

If you are always creating unique files names ... fine.

Then it's a matter of figuring out how to make sure that "each" unique file name created is available everywhere on the cluster ... (if you need to do that).

If you have software that provides a clustered file system across all of the nodes in the cluster (sorry I forgot your environment) peachy then use a directory that's part of the clustered file system.

Otherwise, NFS mounts may be a solution.

Of course you can always create them "almost anywhere" and have some kind of script that periodically runs to find them anywhere and collect them into one spot. Received on Tue Nov 23 2004 - 07:52:25 CST