Re: Unable to shutdown the database

From: joel garry <joel-garry_at_home.com>
Date: Fri, 2 Nov 2012 11:39:07 -0700 (PDT)
Message-ID: <b7da7dff-7fd6-489b-95c1-3c3d269bd9ad_at_b9g2000pba.googlegroups.com>

On Nov 2, 10:13 am, Robert Klemme <shortcut..._at_googlemail.com> wrote:
> On 02.11.2012 15:53, epoko..._at_gmail.com wrote:
>
> > Perhaps it was "stuck" doing the "canceling, etc." for some odd
> > reason. 30 minutes for a shutdown is not normal. Makes even less
> > sense since it appears your startup was successful.
>
> Couldn't it also be the instance recovery triggered by the "startup"
> after "shutdown abort" not having been finished?
>
> Cheers
>
>         robert
>
> --
> remember.guy do |as, often| as.you_can - without endhttp://blog.rubybestpractices.com/

That does make perfect, obvious sense. Also, someone answered this in the forums with a pointer to MOS 1194229.1, which notes a memory configuration shortage can do this. I've seen memory leaks in dbconsole java processes on hp-ux that seem to add up over a couple to 6 months (and in older versions, memory fragmentation from the i/o usage by rman). dbconsole alerts me to excessive memory usage when OMS decides to do stuff _after_ unusually heavy usage of the db is over, D'Oh!

jg

--
_at_home.com is bogus.
Republicans + databases = http://www.utsandiego.com/news/2012/oct/31/mailer-accusing-voters-upsets-some/

Received on Fri Nov 02 2012 - 19:39:07 CET