This is an archived copy of the Xen.org mailing list, which we have preserved to ensure that existing links to archives are not broken. The live archive, which contains the latest emails, can be found at http://lists.xen.org/
Home Products Support Community News


[Xen-users] Re: System reboots after "Scrubbing free RAM" Xen 4.0.1 linu

To: xen-users@xxxxxxxxxxxxxxxxxxx
Subject: [Xen-users] Re: System reboots after "Scrubbing free RAM" Xen 4.0.1 linux 2.6.37-rc3
From: Marek Kaszycki <Marek.Kaszycki@xxxxxxx>
Date: Thu, 25 Nov 2010 23:43:50 +0100
Delivery-date: Fri, 26 Nov 2010 03:28:26 -0800
Envelope-to: www-data@xxxxxxxxxxxxxxxxxxx
List-help: <mailto:xen-users-request@lists.xensource.com?subject=help>
List-id: Xen user discussion <xen-users.lists.xensource.com>
List-post: <mailto:xen-users@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/mailman/listinfo/xen-users>, <mailto:xen-users-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/mailman/listinfo/xen-users>, <mailto:xen-users-request@lists.xensource.com?subject=unsubscribe>
Organization: Sun Microsystems
Sender: xen-users-bounces@xxxxxxxxxxxxxxxxxxx
User-agent: Opera Mail/10.63 (Linux)
Okay, there is some breakthrough...

1. I've started the system with 5 GB allocated for dom0. No go.
2. Next I though, what the heck, and assigned 6 GB, no go.
3. Finally decided to assign 8 GB, no go, but another message appeared, namely that Xen was unable to allocate I/O for IOMMU, Xen gave this information and informed that it's rebooting in 5 seconds. I thought this may be the key.
4. For the next try, I disabled IOMMU and assigned 5 GB to dom0. No go.
5. In the final try, I assigned 8 GB to dom0, no go, but Xen also didn't stop and did not inform me that it's rebooting in 5 seconds, just loaded the kernel.

Apparently I'm not hitting the same bug as Quimi.

Or maybe I am. I'm thinking now if maybe the following .config option is wrong?
Perhaps Xen is expecting a significant fraction of this amount of memory to be available for dom0 (half? quarter? one-eight?) and since it's not getting the expected amount, it simply reboots? This would be rather unexpected, since the kernel is not loading at all, but could it be possible? I'd be grateful if someone provided an explanation of that kernel config option.
The squeeze 2.6.32 kernel has the memory set to 32.
I'll recompile the kernel with max domain memory set to 32 and, if that fails, to 16, just not tonight (I'll get to it tomorrow).

Just for the record, since no one asked, I figure it's not significant, but I am running the Xen hypervisor as provided by squeeze (4.0.1-1), there wouldn't be any caveats or catches against running that version, would there?


Xen-users mailing list

<Prev in Thread] Current Thread [Next in Thread>