Thanks for the quick answer.

Sorry, but I was unable to reproduce the crash, as I reinstalled the
test server. I suspect this to be linked in some kind with tap:aio.
Maybe I forget to blacklist in domU or dom0, the first time. IO seems to
be really slower than with file on an older kernel.

I still get this one
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/172632 but I'm
unable to get any stack trace, except I could pg-up sometime and see the
same message as pasted.

I finally used Etch with 2.6.18 kernel, used file instead of tap:aio (note that 
I had to patch hardy to generate proper xen config file (with tap:aio  inside))
I backported Lenny xen 3.2 stuffs and it works like a charm without patching 
anything.
Looks like xen is far from stable on recent kernel.
Hardy domU works very well on this Debian/Etch + Xen 3.2 and all the annoying 
message about libc6 and tls deseapeared.

I was a bit deceived not to be able to run xen easily on a fresh Hardy using 
xen-tools. 
Maybe a better idea to use an other virtualisation solution on Ubuntu or not to 
use tap:aio (disabled in Etch Kernel), any suggestion?

-- 
kernel crashed system freeze
https://bugs.launchpad.net/bugs/229033
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.

-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to