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


Re: [Xen-devel] Trouble with Xen 2.0

To: Jérôme Petazzoni <jp@xxxxxxxx>
Subject: Re: [Xen-devel] Trouble with Xen 2.0
From: Keir Fraser <Keir.Fraser@xxxxxxxxxxxx>
Date: Mon, 29 Nov 2004 23:43:52 +0000
Cc: Xen-devel@xxxxxxxxxxxxxxxxxxxxx
Delivery-date: Mon, 29 Nov 2004 23:44:19 +0000
Envelope-to: xen+James.Bulpin@xxxxxxxxxxxx
In-reply-to: Your message of "Mon, 29 Nov 2004 16:02:41 +0100." <41AB3A11.8090109@xxxxxxxx>
List-archive: <http://sourceforge.net/mailarchive/forum.php?forum=xen-devel>
List-help: <mailto:xen-devel-request@lists.sourceforge.net?subject=help>
List-id: List for Xen developers <xen-devel.lists.sourceforge.net>
List-post: <mailto:xen-devel@lists.sourceforge.net>
List-subscribe: <https://lists.sourceforge.net/lists/listinfo/xen-devel>, <mailto:xen-devel-request@lists.sourceforge.net?subject=subscribe>
List-unsubscribe: <https://lists.sourceforge.net/lists/listinfo/xen-devel>, <mailto:xen-devel-request@lists.sourceforge.net?subject=unsubscribe>
Sender: xen-devel-admin@xxxxxxxxxxxxxxxxxxxxx
> I tested Xen 2.0 and basic behaviour looks solid and stable, but 
> "dynamic" features didn't work at all for me. More specifically ; with 
> the precompiled build :
> - saving a domain on a P4 computer was erratic (the save process was 
> hung ; dom0 and the to-be-saved VM were still running fine) ; it worked 
> "sometimes" (with no apparent influence of memory load, cpu load, etc). 
> The logs showed the line indicating a save, but nothing more (nothing 
> more than when the save succeeds)

I tested save/restore a little myself recently and couldn't get it to
crash. But various people have seen the crash in time_resume(). Maybe
if I could have a crash dump + vmlinux image file I might be able to
work out what's going on. It'd be even better if I could reproduce the
problem though.

> - using the balloon interface to reduce the memory worked, but reducing 
> it by a large increment caused the affected VM to crash, spitting 
> repeatedly the same message over the console (I didn't have means to 
> save those messages yet, sorry, but I'll try to get them asap)

The balloon driver needs some cleaning and better integration with all
the places in the kernel that increase/decrease the domain's memory

> - OTOH, we could migrate a linux VM from a P4 machine to a Celeron one, 
> and it kept working for a while (but /proc/cpuinfo showed wrong flags, 
> so maybe some optimized code would have crashed)

That just sounds like a bad idea. Suspend/resume or migrating across
very different hardware configurations is asking for trouble. It's
intended for canning images or transferring workloads across a
homegeneous cluster. In particular, 'downgrading' to a lesser CPU is
asking for trouble -- e.g., software RAID trying to use non-existent
SSE instructions, to pull a random example out of the air.

 -- Keir

SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now. 
Xen-devel mailing list

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