 | | | "Out of Memory: Killed process " errors on server running Oracle | "Out of Memory: Killed process " errors on server running Oracle 2007-07-19 - By Gene
Back
We're still weighing the options on our end. We plan on upgrading to the
update 9 kernel and seeing if that resolves anything. Also we're adding
another server to the cluster to see if that helps alleviate the problem.
I'll post more info as we get farther along.
-Gene
On 7/17/07, Eric Sisler <esisler@(protected)> wrote:
>
> Hi Gene,
>
> > Every 5-6 days I'm getting "Out of Memory: Killed process" messages
> > on a Redhat AS 3 server running Oracle 9.20. These errors will occur
> > for a couple hours and will eventually lead to Oracle shutting down.
> > The load on the server did increase recently, but there's still 1.8GB
> > memory free on the system during these errors.
> >
> > The box is running the smp kernel. Should I do something like
> > increase the stack size or use the hugemem kernel?
>
> Were you able to find a fix or workaround to this problem? I'm seeing
> the same thing on one of my VMware GSX/Server hosts. This is apparently
> related to the out of memory process killer on 2.6.x kernel series - in
> some cases it appears to be a little too aggressive when memory gets
> low. The behavior of the OOM process killer also changed between the
> 2.4.x & 2.6.x kernels. From a VMware KB article:
>
> "...This problem results from a change in the way the Linux version 2.6
> kernel's OOM process killer works. In the version 2.4 kernel, the OOM
> process killer killed the process that requested more memory than could
> be provided. In version 2.6, this behavior changed so that the OOM
> process killer kills the largest active process, regardless of the
> process whose request exceeded available memory."
>
> In my case it's the VM with the largest virtual memory that get's killed
> when resources get tight.
>
> I'm still researching possible solutions.
>
> -Eric
>
> --
> Eric Sisler <esisler@(protected)>
> Library Network Specialist
> Westminster Public Library
> Westminster, CO USA
>
> Linux - Don't fear the Penguin.
> Want to know what we use Linux for?
> Visit http://wallace.westminster.lib.co.us/linux
>
> --
> Taroon-list mailing list
> Taroon-list@(protected)
> https://www.redhat.com/mailman/listinfo/taroon-list
>
We're still weighing the options on our end. We plan on upgrading to
the update 9 kernel and seeing if that resolves anything. Also we're
adding another server to the cluster to see if that helps alleviate the problem
.
<br><br>I'll post more info as we get farther along.<br><br>-Gene<br><br>
<br><div><span class="gmail_quote">On 7/17/07, <b class="gmail_sendername">Eric
Sisler</b> <<a href="mailto:esisler@(protected)">esisler
@(protected)
</a>> wrote:</span><blockquote class="gmail_quote" style="border-left: 1px
solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">Hi
Gene,<br><br>> Every 5-6 days I'm getting "Out of
Memory: Killed process" messages
<br>> on a Redhat AS 3 server running Oracle 9.20. These errors
will occur<br>> for a couple hours and will eventually lead to Oracle
shutting down.<br>> The load on the server did increase recently, but there&
#39;s still
1.8GB<br>> memory free on the system during these errors.<br>><br>>
The box is running the smp kernel. Should I do something
like<br>> increase the stack size or use the hugemem kernel?<br><br>Were you
able to find a fix or workaround to this problem? I'm seeing
<br>the same thing on one of my VMware GSX/Server hosts. This is
apparently<br>related to the out of memory process killer on 2.6.x kernel
series - in<br>some cases it appears to be a little too aggressive when memory
gets
<br>low. The behavior of the OOM process killer also changed between
the<br>2.4.x & 2.6.x kernels. From a VMware KB article:<br><br>
"...This problem results from a change in the way the Linux version 2.6<br
>kernel's OOM process killer works. In the version
2.4 kernel, the OOM<br>process killer killed the process that requested more
memory than could<br>be provided. In version 2.6, this behavior changed so that
the OOM<br>process killer kills the largest active process, regardless of the
<br>process whose request exceeded available memory."<br><br>In my case it
's the VM with the largest virtual memory that get's killed<br>when
resources get tight.<br><br>I'm still researching possible solutions.
<br><br>-Eric<br><br>--<br>Eric Sisler <<a href="mailto:esisler@(protected)
.lib.co.us">esisler@(protected)</a>><br>Library Network Specialist
<br>Westminster Public Library<br>Westminster, CO USA<br><br>Linux - Don't
fear the Penguin.
<br>Want to know what we use Linux for?<br>Visit <a href="http://wallace
.westminster.lib.co.us/linux">http://wallace.westminster.lib.co.us/linux</a><br>
<br>--<br>Taroon-list mailing list<br><a href="mailto:Taroon-list@(protected)">
Taroon-list@(protected)</a><br><a href="https://www.redhat.com/mailman/listinfo
/taroon-list">https://www.redhat.com/mailman/listinfo/taroon-list</a><br><
/blockquote></div><br>
--
Taroon-list mailing list
Taroon-list@(protected)
https://www.redhat.com/mailman/listinfo/taroon-list
|
|
|