[Xen-bugs] [Bug 743] New: Unable to start more than a few domU's (Backend device not found)

[bugzilla-daemon@xxxxxxxxxxxxxxxxxxx]

http://bugzilla.xensource.com/bugzilla/show_bug.cgi?id=743

           Summary: Unable to start more than a few domU's  (Backend device
                    not found)
           Product: Xen
           Version: 3.0.2
          Platform: x86-64
        OS/Version: All
            Status: NEW
          Severity: major
          Priority: P2
         Component: Guest-OS
        AssignedTo: xen-bugs@xxxxxxxxxxxxxxxxxxx
        ReportedBy: tejasvia@xxxxxxxxx


(Before somebody dismisses this as a problem with not enough loop devices, let
me clarify that this not related to max_loop=255 for the loop device)

I am running xen 3.0.2-2 compiled from sources on a RHEL4 amd64 box. After I
start "n" number of VM's, I am unable to start any more and they fail with

# xm create vm38-config
Using config file "vm38-config".
Error: Device 771 (vbd) could not be connected. Backend device not found

I am using filesystem images mounted through the loopback device. The number of
VMs that can start, "n",  seems to depend on the number of vbd's per VM. If I
have 2 VBD's per VM (hda1/swap), I am able to start around 15 VM's and with 3
vbd's (hda1/2/swap), around 10 can be started and the number that I can start
is not consistent. There was one instance when I restarted the machine and was
ablt to start upto 40 VM's (with 3 vbd's). I am unable to get that working
again.

If the failure happens during the creation of the initial few VM's, then a
retry will cause it to pass! I have my loopback module to start with 255
loopback devices. They are created fine by "udev",

Looking at the logs and enabling hotplug debugging, I was able to identify that
the create script was erroring out when waiting for the (hotplug notification?)
vbd device (DevController.py).  The error could either be the one above or the
one saying "hotplug scripts are not working". I also increased the timeout
value from 10 to 100 in the script.

I have two machines of RHEL4 amd64 setup with xen, and one has a gig of RAM and
the other 8G of RAM. Both machines exhibit this behavior. I started with a
configuration of 100MB per VM and then gradually shifted down to 25MB per VM,
and the error still happens.

Digging for answers, I hit this mail from a person facing a similar problem:
http://lists.xensource.com/archives/html/xen-users/2006-07/msg00496.html

My xend-config.sxp is unmodifed and defaults to dom0-min-mem of 196.


-- 
Configure bugmail: 
http://bugzilla.xensource.com/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

_______________________________________________
Xen-bugs mailing list
Xen-bugs@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-bugs