This a boring BUG which has a  long history,  from xen 3.* ~ 4.4rc2, which is always occurred when guest os is centos 6+,   as i know.
But it is very strange that only a few guys mention  it and care it.
please read following links:
http://lists.xen.org/archives/html/xen-devel/2013-08/msg01315.html 
http://xen.1045712.n5.nabble.com/xm-save-checkpoint-problem-td2519789.html 


  




herbert cland

From: Don Slutz
Date: 2014-01-25 03:06
To: Konrad Rzeszutek Wilk; Don Slutz
CC: xen-devel@lists.xen.org
Subject: Re: [Xen-devel] [TESTDAY] Test report
On 01/24/14 11:43, Don Slutz wrote:

On 01/24/14 09:58, Konrad Rzeszutek Wilk wrote:

On Thu, Jan 23, 2014 at 05:03:39PM -0500, Don Slutz wrote:

On 01/20/14 18:35, Konrad Rzeszutek Wilk wrote:

Don Slutz <dslutz@verizon.com> wrote:

[snip]


WARNING: g.e. still in use!
WARNING: g.e. still in use!
WARNING: g.e. still in use!
pm_op(): platform_pm_resume+0x0/0x50 returns -19
PM: Device i8042 failed to resume: error -19
INFO: task sadc:22164 blocked for more then 120 seconds.
"echo 0 >..."
INFO: task sadc:22164 blocked for more then 120 seconds.

[root@dcs-xen-54 ~]# xl des 17
[root@dcs-xen-54 ~]# xl restore -V
/big/xl-save/centos-6.4-x86_64.0.save


Not sure if this is expected or not.

I think Ian saw this with the 'fast-cancel' something resume but I might be incorrect. Did it work if you used xend (you might have to configure it be enabled)?

I have not used xend/xe in a long time.  I did need to configure it.

Does not start:


# /etc/init.d/xend start
WARNING: Enabling the xend toolstack.
xend is deprecated and scheduled for removal. Please migrate
to another toolstack ASAP.
Traceback (most recent call last):
  File "/usr/sbin/xend", line 110, in <module>
    sys.exit(main())
  File "/usr/sbin/xend", line 91, in main
    start_blktapctrl()
  File "/usr/sbin/xend", line 77, in start_blktapctrl
    start_daemon("blktapctrl", "")
  File "/usr/sbin/xend", line 74, in start_daemon
    os.execvp(daemon, (daemon,) + args)
  File "/usr/lib64/python2.7/os.py", line 344, in execvp
    _execvpe(file, args)
  File "/usr/lib64/python2.7/os.py", line 380, in _execvpe
    func(fullname, *argrest)
OSError: [Errno 2] No such file or directory


How important is it to try this?

It tells us whether the issue is indeed with the 'fast-cancel' thing.

But, I do recall seeing a patch from Ian Jackson for this - I just
don't remember what it was called - it was posted here and perhaps
applying it would help?


I have not found a patch.  The bug #30:


http://bugs.xenproject.org/xen/bug/30

Looks to me like the issue I am seeing.  The Guest I was testing this on is an old kernel (as far as this bug) and so I would expect it might be related.


Here is what I see that may link it to this:


From: Ian Campbell <Ian.Campbell@citrix.com>
To: Ian Jackson <Ian.Jackson@eu.citrix.com>
Cc: xen-devel@lists.xen.org
Subject: Re: [Xen-devel] 3.4.70+ kernel WARNING spew dysfunction on failed migration
Date: Fri, 10 Jan 2014 10:26:31 +0000


...

Looks like RHEL4 (linux-2.6.9-89.0.16.EL kernel) doesn't have the
support for the new mode at all.

It would probably be wise to validate this under xend before chasing
red-herrings with xl.

Ian.


So I will continue the fight to get xend running.

   -Don




Ah, now have xend running, but still need to convert from an xl.cfg file to a xmdomain.cfg...  Having not used xm in years, this will take a while.  The man page does not say how to build an hvm guest.

I did another test and Fedora 19 (x86_64) saved just fine.  So this looks to be bug #30.

If I understand this all, this means that the older linux kernels will fail this way also if a migrate fails and the source guest is restarted.

So, do I continue to get xend working, or just say it is an example of bug #30?

    -Don Slutz







    -Don Slutz