Re: Quickest Way To Recreate a Database
From: Michael Wehrle <michaelw436_at_gmail.com>
Date: Tue, 30 Jun 2009 20:34:57 -0400
Message-ID: <54b1f9980906301734v13cb1b58oc7651b9bf81e542d_at_mail.gmail.com>
>
> Mathias, I am seeing unencrypted customer data that used to be unencrypted,
> but is now encrypted in the system tablespace, and vice versa, sys segment
> data in all datafiles. I believe a new thread has begun documenting how or
> even why that occurs named "Ghost Data" but it is simply reproducible.
>
> We have considered online redefinition, and that would work, except for it
> doesn't work on the system tablespace. We may still be able to use it in a
> multistep strategy of sorts.
>
> You are right about the data living on the physical disk long after it
> doesn't logically belong to the database. For that, we are using EMC storage
> array currently, which I hardly know in detail, just in concept. I will have
> to rely on our system admins to clean the physical disks properly.
>
>
Date: Tue, 30 Jun 2009 20:34:57 -0400
Message-ID: <54b1f9980906301734v13cb1b58oc7651b9bf81e542d_at_mail.gmail.com>
>
> Mathias, I am seeing unencrypted customer data that used to be unencrypted,
> but is now encrypted in the system tablespace, and vice versa, sys segment
> data in all datafiles. I believe a new thread has begun documenting how or
> even why that occurs named "Ghost Data" but it is simply reproducible.
>
> We have considered online redefinition, and that would work, except for it
> doesn't work on the system tablespace. We may still be able to use it in a
> multistep strategy of sorts.
>
> You are right about the data living on the physical disk long after it
> doesn't logically belong to the database. For that, we are using EMC storage
> array currently, which I hardly know in detail, just in concept. I will have
> to rely on our system admins to clean the physical disks properly.
>
>
-- http://www.freelists.org/webpage/oracle-lReceived on Tue Jun 30 2009 - 19:34:57 CDT