Hi all
I'm working with Xen, openSuse 11.1 and OpenTC for my thesis project.
I recently update the xen-tools-vgallium-dom0, kernel-xen,
kernel-xen-base and kernel-xen-extra rpms to the newest versions:
- kernel 2.6.27.90-135.1.i586
- xen-tools 3.3.1_18494_03-4.2-i586
and something wrong appened in my openSuse 11.1 distro.
1)
If I start compositor with the script:
--------------------------------------------------------------------
#!/bin/bash
modprobe gntdev
EGL_DRIVER=librawgal /*the-xen3d-path*/xen3d --trusted-domain $@
--------------------------------------------------------------------
and than start my domU (based on xen kernel 2.6.27.90-135.1.i586, as
dom0) with the command:
- xm create domU
xm prompts this error:
--------------------------------------------------------------------
Using config file "/etc/xen/vm/domU".
Xend has probably crashed! Invalid or missing HTTP status code.
--------------------------------------------------------------------
The command:
- xm list
shows this lines:
--------------------------------------------------------------------
Name ID Mem VCPUs State Time(s)
Domain-0 0 992 2 r----- 50.4
Domain-Unnamed 1 1024 1 --p--- 0.0
--------------------------------------------------------------------
So, I can only do this:
- xm destroy Domain-Unnamed
2)
If I don't start compositor and I only start domU in a console with the
following commands:
- xm create domU
- xm console domU
domU starts normally, but, after the login, when I try to shutdown the
domain (always in console, no X and no compositor) with the command:
- shutdown -h now
the domain crashes at the end of the shutdown procedure.
This is what I see on the console:
--------------------------------------------------------------------
INIT: Switching to runlevel: 0
INIT: Sending processes the TERM signal
INIT: Sending processes the KILL signal
/etc/init.d/rc: line 88: 0: command not found
Boot logging started on /dev/ttyS0(/dev/console) at Fri Jun 12 09:44:15 2009
Master Resource Control: previous runlevel: S, switching to runlevel: 0
Running /etc/init.d/halt.local done
Serial: 8250/16550 driver8 ports, IRQ sharing enabled
modprobe: FATAL: Error inserting 8250
(/lib/modules/2.6.27.90-135-xen/kernel/drivers/serial/8250.ko): Device
or resource busy
Sending all processes the TERM signal... done
Sending all processes the KILL signal... done
The system will be halted immediately.
md: stopping all md devices.
xen console-0: xenbus_dev_shutdown: device/console/0: Initialising !=
Connected, skipping
vfb vfb-0: xenbus_dev_shutdown: device/vfb/0 timeout closing device
System halted.
--------------------------------------------------------------------
After this messages, the console don't turn back to the command prompt,
but it stops working.
Any following "xm" command (xm list, xm destroy domU...) in a different
console fails (unable to connect to xend, it ask me if xend is still
running).
But xend IS still running! In the processes list, there are 2 xend
processes, one of wich that runs at 50% of cpu (???) even if there are
no more running domains.
If I try to stop xend with the command:
- /etc/init.d/xend stop
it stops working, but when I try to restart it with:
- /etc/init.d/xend start
the 2 processes restart and the second one restarts to work at 50% of cpu.
Before installing the new rpm's, I had no problems (with xen kernel
2.6.27.90-105 and the previous version of xen-tools-vgallium-dom0).
What I can do to get my openSuse and domU's working like before?
And, if it shouldn't be possible to repair it...
Unfortunately, I can't roll back to the previous rpm's, because they are
no more in the OpenTC repository.
It's possible to find those old rpm's? I have to continue working on my
thesis project and I can't wait a new rpm's release.
Thank you very much in advance
Best regards
_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users
|
Home