RE: [Xen-devel] problems starting domains

[James Harper <JamesH@xxxxxxxxxxxxxxxx>]

hmmm.... sda7 definitely contains a valid filesystem. It was actually running under uml.

 

is there any way I can see the output? any hints about why it might exit?

 

here's the guts of my config file, maybe you can see something wrong with it...

 

import sys
import xenctl.ip
try:
    vmid = int(vmid) # convert to integer
except:
    raise ValueError, "Variable 'vmid' must be an integer"
if vmid <= 0:
    raise ValueError, "Variable 'vmid' must be greater than 0"
kernel = "/boot/vmlinuz-2.4.26-xenU"
memory = 128
name = "sbssgw1 vm %d" % vmid
cpu = vmid  # set based on vmid (mod number of CPUs)
nics=1
mac = [ "aa:00:00:00:00:11" ]
hostname= "vm%d" % vmid
disk = [ 'phy:sda7,sda7,w' ]
root = "/dev/sda7 ro"
extra = "4 VMID=%d " % vmid

/boot in domain0 is a different partition to /... that wouldn't make a difference would it?

 

thanks

 

James

---

From: Ian Pratt
Sent: Fri 18/06/2004 10:58 PM
To: James Harper
Cc: xen-devel@xxxxxxxxxxxxxxxxxxxxx; Ian.Pratt@xxxxxxxxxxxx
Subject: Re: [Xen-devel] problems starting domains

> I'm running the latest bk source, and am having trouble starting domains. Domain0 boots okay though.
> 
> with xend putting output to stdout, starting a domain gives me this:

The new xend is pretty chatty right now, but having the verbose
output is useful until the bugs get shaken out...

I suspect what is happening is that the new guest OS kernel is
exiting for some reason. The most likely reason is that it can't
mount its root partition. I assume your sda7 contains a valid
root file system i.e. you can mount sda7 from domain0 OK?
[NB: make sure the filesystem is unmounted before trying to mount
it from another domain!]

Normally, the control tools would let you see the console output
from the domain and you could see what's going on. It looks like
you're hitting multiple bugs in the new tools -- Mike is best
placed to try and interpret the various back traces as I don't
know my way around the new code yet.

BTW: Thanks for bearing with us while we complete this major bit
of integration work to the xeno-unstable tree. I think things
will have settled down substantially by the end of next week,
proving you guys keep beating on it and telling us about the bugs
;-)

Ian