Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
![]() |
![]() |
Home -> Community -> Usenet -> c.d.o.server -> Re: Migrate from DMT to LMT when fet$ and uet$ counts > 700,000
Charles wrote:
>
> All,
>
> I have a 1.8TB database with all Dictionary Managed Tablespaces.
>
> It's a very active SAP Business Warehouse application.
>
> We've sort of reached 'critical mass' with the fragmentation of it,
> and now SMON coalescing is so slow that "Disk Space Transaction" locks
> are hanging up the entire db.
>
> I've added events to disable SMON coalescing while we deal with this.
>
> Oracle is telling me I must run tablespace coalescing at 'quiet' times
> prior to migrating the tablespace to Locally Managed Tablespaces.
>
> I want to migrate them now to eliminate the contention on the fet$ and
> uet$ tables in the SYSTEM tablespace.
>
> I await Oracle's technical reasons why I cannot migrate now.
>
> I have done the migration in all non-production instances of this
> warehouse application; they ran seemingly ok and those databases are
> performing well, according to the warehouse users.
>
> What is your experience with this? What's the problem or the bad
> results in the LMT if I migrate these tablespaces now?
>
> Am looking for the quick fix now, to be following to reorgs later.
>
> Thank you.
>
> Charles
Basically you're damned if you do, damned if you don't.
Whilst migrating to LMT as is will leave you with a soup of extent sizes, it should still be better than where you are now.
You may want to consider
Convincing management to do (b) once (a) is done might be a challenge...
C
-- Connor McDonald Co-author: "Mastering Oracle PL/SQL - Practical Solutions" ISBN: 1590592174 web: http://www.oracledba.co.uk web: http://www.oaktable.net email: connor_mcdonald_at_yahoo.com Coming Soon! "Oracle Insight - Tales of the OakTable" "GIVE a man a fish and he will eat for a day. But TEACH him how to fish, and...he will sit in a boat and drink beer all day" ------------------------------------------------------------Received on Fri Oct 22 2004 - 08:57:57 CDT
![]() |
![]() |