Re: [Xen-devel] xen page fault

[Keir Fraser <Keir.Fraser@xxxxxxxxxxxx>]

> 
> In the process of sorting out my ganglia setup, I faulted two xen's,
> one of which just happened to be on a serial line.
> The fault occured when I shutdown the xenU domains on 16 hosts.  The two
> hosts that crashed are the two to which gmetad linked from a host
> outside the cluster.   So I am guessing some funny business with multicast
> pkts on a xenU that did _not_ have IP_MULTICAST configured.   In fact, I
> was rebooting to start a xenU that did have MULICAST configured.
> xen0 does not have MULTICAST configured either but it does have
> CONFIG_XEN_WRITABLE_PAGETABLES on.
> 
> Have people run gmond on xenU as a data_source for gmetad?

This is unlikely to have anything to do with networking or
multicast. Xen's per-page reference counts ahve got out of sync, such
that a page has ended up freed whil eit still has a non-zeo 'type
count'. This may be related to the crashes doing decrease_reservation
in the net frontend driver. Look slike at least on epage-management
bug lives on... :-(

 -- Keir


-------------------------------------------------------
This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
Use IT products in your business? Tell us what you think of them. Give us
Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out more
http://productguide.itmanagersjournal.com/guidepromo.tmpl
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxxx
https://lists.sourceforge.net/lists/listinfo/xen-devel