Re: Upgrading Grid Infra for Standalone ASM from to

From: Don Seiler <>
Date: Fri, 15 Apr 2011 13:53:24 -0500
Message-ID: <>

FYI if anyone is interested. The workaround (removing the old grid home directory and restarting) worked and core dumping stopped. Pushing it back to Oracle to get the bug re-opened or a new bug created. In just over 1 week it had created 8.2Gb of these core dump files.

Thanks for the response, yes in the meantime I planned to just create a regular script to run a few times a day and remove these core dump files.


On Thu, Apr 14, 2011 at 9:28 AM, Niall Litchfield <> wrote:

> Not that I'm bored waiting for long running processes to do their stuff or
> anything...
> Anyway I've done a wee bit of research for you and determined the
> following.
> 1) You are not alone - there's at least one thread on the same issue.
> 2) Diskmon is an exadata feature! that on most platforms is disabled in the
> absence of exadata at least in 11.1
> 3) in 11.1 you could make a change to init.cssd to stop the diskmon process
> starting, you can't on 11.2.
> 4) some patches for GI were delivered with a missing diskmon.bin though
> that seems to be an 11.1 problem.
> It's likely therefore that you may be able to disable the diskmon process
> within CRS. Clearly this is not something I would do without
> a) talking to support given the almost complete lack of documentation for
> the diskmon background process (3 mentions in 11.2 doc set, none in 11.1 :)
> )
> b) experimenting carefully on a test system.
> On Wed, Apr 13, 2011 at 5:00 PM, Don Seiler <> wrote:
>> Wondering if anyone else out there has done upgrades of Grid
>> Infrastructure for standalone ASM from to I seem to be
>> encountering a new bug every time I do it.
>> The latest one results in core dump files being created every 10 minutes
>> for no reason. This is bug 10283819, although it's been closed as "not a
>> bug". The original poster said that the only workaround he found was to
>> remove the old grid home directory (which I'd rather not do until
>> I'm well into running on and restart all services. So far in just
>> over 5.5 days, I have over 6.5GB of these core dump files in the
>> $GRID_HOME/log/<hostname>/diskmon/ directory.
>> If anyone else has done such an upgrade, I'd like to know if you see this
>> as well. Definitely seems like it's not "not a bug" to me.
>> --
>> Don Seiler
> --
> Niall Litchfield
> Oracle DBA

Don Seiler

Received on Fri Apr 15 2011 - 13:53:24 CDT

Original text of this message